Remotesensing 13 00705 v3

remote sensing
Article
Canopy Top, Height and Photosynthetic Pigment Estimation
Using Parrot Sequoia Multispectral Imagery and the Unmanned
Aerial Vehicle (UAV)
Veronika Kopačková-Strnadová 1, * , Lucie Koucká 1 , Jan Jelének 1 , Zuzana Lhotáková 2 and Filip Oulehle 1
1 Czech Geological Survey, Klárov 3, 11821 Prague, Czech Republic; lucie.koucka@geology.cz (L.K.);
jan.jelenek@geology.cz (J.J.); filip.oulehle@geology.cz (F.O.)
2 Department of Experimental Plant Biology, Faculty of Science, Charles University, Viničná 5,
12844 Prague, Czech Republic; zuzana.lhotakova@natur.cuni.cz
* Correspondence: veronika.kopackova@seznam.cz; Tel.: +420-257-089-481
Abstract: Remote sensing is one of the modern methods that have significantly developed over the
last two decades and, nowadays, it provides a new means for forest monitoring. High spatial and
temporal resolutions are demanded for the accurate and timely monitoring of forests. In this study,
multi-spectral Unmanned Aerial Vehicle (UAV) images were used to estimate canopy parameters
(definition of crown extent, top, and height, as well as photosynthetic pigment contents). The UAV
images in Green, Red, Red-Edge, and Near infrared (NIR) bands were acquired by Parrot Sequoia
camera over selected sites in two small catchments (Czech Republic) covered dominantly by Norway
spruce monocultures. Individual tree extents, together with tree tops and heights, were derived

from the Canopy Height Model (CHM). In addition, the following were tested: (i) to what extent can
Citation: Kopačková-Strnadová, V.; the linear relationship be established between selected vegetation indexes (Normalized Difference
Koucká, L.; Jelének, J.; Lhotáková, Z.; Vegetation Index (NDVI) and NDVIred edge ) derived for individual trees and the corresponding
Oulehle, F. Canopy Top, Height and
ground truth (e.g., biochemically assessed needle photosynthetic pigment contents) and (ii) whether
Photosynthetic Pigment Estimation
needle age selection as a ground truth and crown light conditions affect the validity of linear models.
Using Parrot Sequoia Multispectral
The results of the conducted statistical analysis show that the two vegetation indexes (NDVI and
Imagery and the Unmanned Aerial
NDVIred edge ) tested here have the potential to assess photosynthetic pigments in Norway spruce
Vehicle (UAV). Remote Sens. 2021, 13,
705. https://doi.org/10.3390/
forests at a semi-quantitative level; however, the needle-age selection as a ground truth was revealed
rs13040705 to be a very important factor. The only usable results were obtained for linear models when using
the second year needle pigment contents as a ground truth. On the other hand, the illumination
Academic Editor: Luke Wallace conditions of the crown proved to have very little effect on the model’s validity. No study was found
to directly compare these results conducted on coniferous forest stands. This shows that there is a
Received: 12 January 2021 further need for studies dealing with a quantitative estimation of the biochemical variables of nature
Accepted: 12 February 2021 coniferous forests when employing spectral data that were acquired by the UAV platform at a very
Published: 16 February 2021 high spatial resolution.
Publisher’s Note: MDPI stays neutral Keywords: UAV; Parrot Sequoia multispectral camera; photosynthetic pigments; Norway spruce;
with regard to jurisdictional claims in
forest; linear models; ground truth; needle age; crown detection
published maps and institutional affil-
iations.
1. Introduction
Forests play a significant role in the Earth’s ecosystems and they contribute greatly
Copyright: © 2021 by the authors.
to reducing adverse climate change impacts. They provide a natural environment for
Licensee MDPI, Basel, Switzerland.
many species of animals and plants, represent a significant carbon sink, and support an
This article is an open access article
effective hydrological cycle. In addition, forests serve as an important source of timber
distributed under the terms and
conditions of the Creative Commons
and other non-wood materials [1,2]. At the beginning of the 1980s forest health and the
Attribution (CC BY) license (https://
sustainability of their ecosystems became a highly discussed topic for politicians, the public,
creativecommons.org/licenses/by/
and scientists due to the aforementioned functions of forests and the increasing level of
4.0/). damage they suffer [3].
Remote Sens. 2021, 13, 705. https://doi.org/10.3390/rs13040705 https://www.mdpi.com/journal/remotesensing

Remote Sens. 2021, 13, 705 2 of 26
In Europe, temperate forests are mainly affected by climate change and air pollu-
tion [4,5]. During the 20th century, the region on the Czech, Polish, and German borders
was influenced by extensive coal mining, which was linked to large emissions of SO2
and NOx from power plants [6,7]. In order to monitor the process of ecosystem recovery
after the reduction in pollution that started in the 1980s and accelerated in the 1990s, the
Geochemical Monitoring network of small catchments (GEOMON) was initiated across
the Czech Republic. Since 1993, when the GEOMON network began, the data that were
collected from observations of these catchments have been used for many studies [7,8]
mainly corresponding to catchment biogeochemistry (e.g., [6,9–13]). The recent study that
was published by Švik et al. [7] supplied the aforementioned research that was based
on field observations using remote sensing methods that have been also employed in
this study.
Remote sensing techniques have been frequently used to study forest areas for multiple
purposes over the last decade. They have been proven to be less costly and time-consuming
alternatives to ground level research [14]. Satellite and aerial imagery have offered an
opportunity to investigate forests at the regional scale, for example, to estimate forest
biomass, monitor forest cover changes, or classify types of biome [15–18]. The use of
airborne multispectral and hyperspectral sensors has led to closer forest observation, such
as the classification of tree species, monitoring forest health, or estimating chlorophyll
content [19–24]. The newly developed Unmanned Aerial Vehicles (UAV) complement the
established remote sensing (RS) methods. UAVs have increased the number of benefits,
such as acquiring extreme high-spatial resolution data, flexibility in usage and over time,
and the capacity to carry various sensors, such as a multispectral camera, to observe
vegetation health [25,26].
Most of the studies use a combination of UAV complemented by a multispectral
sensor to analyse agriculture crops for precision farming (e.g., tomato, vineyard, or wheat
production), where they usually employ Vegetation Indices (VIs), such as the Normalized
Difference Vegetation Index (NDVI), the Green Normalized Difference Vegetation Index
(GNDVI), or the Soil Adjusted Vegetation Index (SAVI), to monitor crop health [27–32].
Authors monitoring vineyards described the use of multispectral and thermal sensors in
combination to obtain additional information about crop water status [33–35].
When assessing the foliage traits of forest trees, it is necessary to take the tree size,
branch geometry, and tree density, respectively, foliage clumping into account [36]. Dealing
with evergreens, particularly conifers, a higher level of complexity emerges as different
needle age generations contribute to the final signal received by the sensor. These might be
the main reasons why, despite the high potential of UAVs, not many studies have dedicated
an analysis of multispectral data for assessing forests. Recently, UAV-based multispectral
sensing has been used for high-throughput monitoring of the photosynthetic activity in a
white spruce (Picea glauca) seedling plantation [37]. A few more studies on forest health
assessment using the UAVs and multispectral cameras have been published so far. Dash
et al. [38] demonstrated the usefulness of such approaches for monitoring physiological
stress in mature plantation trees, even during the early stages of tree stress when using a
non-parametric approach for the qualitative classification. Chianucci et al. [39] used the true
colour images, together with a fixed-wing UAV, to quantify the canopy cover and leaf area
index of beech forest stands. Dash et al. [40] tested the sensitivity of the multispectral image
data time-series that were acquired by the UAV platform and satellite imagery to detect
herbicide-induced stress in a controlled experiment conducted on a mature Pinus radiata
plantation. Other successful qualitative forest classifications have been demonstrated
while using UAVs together with multispectral [41,42] and hyperspectral cameras [43,44]. A
pioneer approach for estimating tree-level attributes and multispectral indices using UAV
in a pine clonal orchard was recently published by Gallardo-Salazar and Pompa-García [45].
All of these studies listed above deal with qualitative forest classifications, so, a need for
quantitative approaches, focusing on forest biochemical variable estimations employing
UAV-based multispectral sensing, is clearly demanded.
Remote Sens. 2021, 13, 705 3 of 26
For such quantitative spectral-based approaches, which employ very high-spatial

resolution data, an accurate definition of the extent of individual trees is a crucial task.
The delineation of a tree crown boundary has been studied before and several approaches
to detect individual trees have been published. Lim et al. [46] applied a segmentation
method to the RGB orthoimage and the Canopy Height Model (CHM) that was obtained
by UAV. A similar approach using multispectral data instead was described in Díaz-Varela
et al. [47]. A popular technique is hydrological terrain analysis–watershed algorithm,
where the inverse CHM is delineated by the watershed algorithm, where catchment basins
represent individual trees and holes substitute tree peaks [15,48–50]. Among the recent
methods are the LiDAR point cloud segmentation [51,52], which achieves highly accurate
results [48,53–57] and also approaches that employ deep learning techniques to detect
individual trees [58,59].
This study tested, if a UAV that is equipped with a multispectral camera can be
employed for photosynthetic pigment estimation in coniferous trees. As already described,
when assessing coniferous forests, such factors as tree size and density affect the validity
of estimations of structural forest parameters [36]. Moreover, as different needle age
generations contribute to the final image pixel reflectance, it is still not clear what ground
truth should be optimally used (e.g., [60]). To fill these gaps, two small catchments (part of
the GEOMON network in Czech Republic) inhabited by mainly mature Norway spruce
monocultures were selected as test sites based on their relatively close spatial proximity,
alongside differences in parent material, which significantly influenced nutrient availability
for forest ecosystems. The images from a Multi-spectral Unmanned Aerial Vehicle (UAV)
were acquired over these two sites, and the following topics were researched:
• An accurate definition of the individual tree extents (crown delineation) and derivation
of other parameters, such as tree top and height using the UAV-based multispectral data.
• Testing if a linear relationship can be established between selected vegetation indices
(NDVI and NDVIred edge ) that are derived for individual trees and the corresponding
ground truth (e.g., biochemically assessed needle photosynthetic pigment contents).
• Testing whether the needle age selection, as ground truth affects the validity of the
linear models.
• Testing if the tree crown light conditions affect the validity of the linear models.
2. Materials and Methods

2.1. Test Sites
Two test sites representing rural mountainous landscapes in the western part of Bo-
hemia were selected—Lysina (LYS) and Pluhův Bor (PLB) (Figure 1). These two catchments
have been part of the European network—GEOMON established in 1993 to assess changes
in precipitation and stream chemistry after reducing pollution in Eastern Europe. The
selected catchments were heavily affected by acid pollution during the 20th century. Nowa-
days, both sites are part of the Slavkov Forest Protected Landscape Area. The vegetation
cover mostly consists of managed Norway spruce monocultures [10], which are situated
around 800 ma.s.l. The main difference between these two sites is the lithology and soil
type as well as the forest age. Tables 1 and 2 provides detailed catchment characteristics.
Remote Sens. 2021, 13, 705 4 of 26
Remote Sens. 2021, 13, 705 4 of 28
Figure 1.
Figure 1. Location
Location of
of the
the test
test sites Pluhův
sites Pluh ův Bor and Lysina
Bor and Lysina in
in the
the Czech
Czech Republic,
Republic, Europe:
Europe: (a)
(a) the
the test
test sites
sites displayed
displayed on
on aa
map of the Czech Republic; (b) maps of the Lysina and Pluhův Bor catchments with tree stands highlighted;
map of the Czech Republic; (b) maps of the Lysina and Pluhův Bor catchments with tree stands highlighted; an Orthophoto an
Orthophoto of the Czech Republic in the background [61].
of the Czech Republic in the background [61].
Table 1. Characteristics of selected catchments: Lysina and Pluhův Bor [10].

Table 1. Characteristics of selected catchments: Lysina and Pluhův Bor [10].
Forest Spruce Non-
Elevation Forest Spruce Broadleaf Non-
Site Bedrock Soils Age (Year Forest Forested
Site Bedrock Soils (mElevation
a. s. l.) Age (Year Forest
Broadleaf
Forest (ha) Forested
(m a. s. l.) Range) (ha) Forest (ha) Area (ha)
Range) (ha) Area (ha)
Cambisols,
LYS
LYS Granite Cambisols, Podzols
Granite 880880 12–53
12–53 27
27 0.0
0.0 0.4
0.4
Podzols
Magnesic Cambisols,
Magnesic
Stagnic–Magnesic
Cambisols,
PLB Serpentinite 755 41–129 18 0.0 4.0
Cambisols,
Stagnic–
Magnesic Gleysols
PLB Serpentinite Magnesic 755 41–129 18 0.0 4.0
Cambisols,
Magnesic
Table 2. Tree ages at the two catchments selected in this study.
Gleysols
Test Site Average Age Min Age Max Age
Table 2. Tree ages at the two catchments
LYS 1K 14 selected in this study.
12 17
LYS 2K 16 15 18
Test
LYS Site
4K Average
47 Age Min44Age Max53Age
LYS
PLB 1K
2K 14
120 12
109 17
129
PLB 2K
LYS 3K 72
16 70
15 74
18
PLB 5K 47 41 50
LYS 4K 47 44 53
PLB 2K 120 109 129
2.2. In-situ
PLBGround
3K Truth 72 70 74
In PLB
this study,
5K three tree stands
47 were selected in the
41 LYS catchment (LYS 1K, LYS 2K,
50
and LYS 4K) and three stands in the PLB catchment (PLB 2K, PLB 3K, and PLB 5K), these
stands
2.2. wereGround
In-situ the object
Truthof previous long-term research, thus the soil conditions were known
as well as tree height information, which was measured in-situ and modelled using the
In this study, three tree stands were selected in the LYS catchment (LYS 1K, LYS 2K,
LiDAR data [62].
and LYS 4K) and three stands in the PLB catchment (PLB 2K, PLB 3K, and PLB 5K),
To follow up with the previous research activities, the same trees described in Lhotáková
these stands were the object of previous long-term research, thus the soil conditions
et al. [60] were selected at each stand (three trees per stand, in total 18 trees). The branch
were known as well as tree height information, which was measured in-situ and
samples were collected by a climber from the sunlit crown part one day before the UAV-
modelled using the LiDAR data [62].
based data were acquired (August 2018). The needle age was identified and three different
To follow up with the previous research activities, the same trees described in
age classes were sampled: first, second, and a mixed sample of fourth year and older
Lhotáková et al. [60]referred
needles (hereinafter were selected
to as 4thatyear
eachfor
stand (three [60].
simplicity) trees The
per needles
stand, in totalcooled
were 18 trees).
and
The branch samples were collected by a climber from the sunlit
immediately transported to the laboratory where they were kept at a stable temperaturecrown part one day
before
at −20 the UAV-based
◦ C until furtherdata were acquired
processing. (August pigments—chlorophyll
Photosynthetic 2018). The needle age was a, bidentified
and total
and three different age classes were sampled: first, second, and a
carotenoids—were extracted in dimethylformamide following the procedures described mixed sample of fourth
in
year and older needles (hereinafter referred to as 4 th year for simplicity) [60]. The needles
detail by Porra et al. [63] and then spectrophotometrically determined [64].
were cooled and immediately transported to the laboratory where they were kept at a
stable temperature at −20° C until further processing. Photosynthetic
pigments—chlorophyll a, b and total carotenoids—were extracted in
Remote Sens. 2021, 13, 705 5 of 28
Remote Sens. 2021, 13, 705 5 of 26

dimethylformamide following the procedures described in detail by Porra et al. [63] and
then spectrophotometrically determined [64].
2.3. UAV
2.3. UAV Data
Data Acquisition
Acquisition
2.3.1. Equipment
2.3.1. Equipment
In
In this
thisexperiment,
experiment, an an unmanned
unmanned aerial
aerial vehicle
vehicle DJI
DJI Phantom
Phantom 44 (SZ
(SZ DJI
DJITechnology
Technology
Co.,
Co., Ltd. [65]) was employed (Figure 2b). It is a widely used quadcopter weighing1380
Ltd. [65]) was employed (Figure 2b). It is a widely used quadcopter weighing 1380gg
with RGB camera in 4K (4096 ×
with an RGB camera in 4K (4096 × 2160 px) [66]. The UAV was complemented by
an 2160 px) [66]. The UAV was complemented by aa
multispectral
multispectralcamera
cameraParrot
ParrotSequoia scanning
Sequoia scanning system (senseFly
system Inc. [67]).
(senseFly The camera
Inc. [67]). was
The camera
specially designed
was specially to support
designed vegetation
to support studies;
vegetation therefore,
studies; besidesbesides
therefore, Red and Green
Red andbands,
Green
there
bands,is athere
bandisplaced in the
a band Red in
placed edge
theregion as well
Red edge as one
region asband
wellinasthe
oneNear
bandinfrared
in the(NIR)
Near
(Table 3).
infrared (NIR) (Table 3).
Figure 2.2. Unmanned

Figure Unmanned Aerial
Aerial Vehicle
Vehicle (UAV)
(UAV) equipment
equipment used
usedin
inthis
thisstudy:
study:(a)
(a)three-dimensional
three-dimensional
(3D) models of the mounts designed for the Parrot Sequoia multispectral camera; (b)(b)
(3D) models of the mounts designed for the Parrot Sequoia multispectral camera; DJIDJI Phantom
Phantom 4
4 quadcopter with attached Parrot Sequoia camera; and, (c) planned flight path in the flylitchi.com
quadcopter with attached Parrot Sequoia camera; and, (c) planned flight path in the flylitchi.com
web tool.
web tool.
Table 3. Overview of a Parrot Sequoia spectral band setting.
Table 3. Overview of a Parrot Sequoia spectral band setting.
Band Name Spectral Range (nm) Central Wavelength (nm)
Band Name Spectral Range (nm) Central Wavelength (nm)
Green 530–570 550
Green
Red 530–570
640–680 550
660
Red 640–680 660
Red edge
Red edge(RE)
(RE) 730–740
730–740 735
735
Near infrared
Near infrared(NIR)
(NIR) 770–810
770–810 790
790
The Parrot Sequoia consists of two parts—the main camera and a sunshine sensor
The Parrot Sequoia consists of two parts—the main camera and a sunshine sensor
calibrating the measured
calibrating the measured spectral
spectral radiation
radiation by
by the
the main
mainsensor.
sensor. These
These two
two parts
parts of
ofthe
the
multispectral camera were attached to the drone by mounts that were designed
multispectral camera were attached to the drone by mounts that were designed by the by the
CGS team
CGS team (Figure
(Figure2a).
2a). The
The 3D
3D models
models ofofmounts
mountswere
werecreated
createdin
inCAD
CADsoftware.
software.ItItwas
was
important to make a free space for the GPS sensor to ensure good signal reception. The
holder for the sunshine sensor had to be placed in a manner, so as not to have any contact
Remote Sens. 2021, 13, 705 6 of 26
with the propellers (Figure 2a). The mounts were printed using the Prusa i3 3D printer
(Prusa Research a.s. [68]).
2.3.2. Data Acquisition

The RGB and Parrot Sequoia multispectral data were obtained for both test sites.
The flights were made on 6 and 7 August 2018, between 11 am and 3 pm to ensure the
multispectral camera captured the required maximum reflected sunlight and eliminate
shadows. Prior to data acquisition, a flight path was planned in the flylitchi.com web
tool (Figure 2c), which was connected to the litchi android application (VC Technology
Ltd. [69]) controlling the UAV. Areas of 40 × 40 m2 were defined to cover the tree groups
and their surroundings. All of the flight paths covering the area of interest were planned in
a north-south direction and in a way so the parallel scanning lines would reach 70% of the
side overlap in order to ensure errorless image mosaicking when creating photogrammetric
products. The flight height was set up according to the highest terrain point (40–70 m above
ground level) and it was 25–30 m above the treetops. The UAV speed was set at 3.6 km/h to
ensure well-focused images were recorded. Sequences of images were recorded every two
seconds (s) with 95% overlap in the direction of the flight. The resulting spatial resolution
of RGB imagery varied between 1.1–1.9 cm/px and 3.76–6.59 cm/px for multispectral
imagery, respectively, according to flight height.
Regarding multispectral data calibration, images of the calibrated reflectance panel
were acquired before and after each flight. The same routine was used for each flight—the
calibration target (Aironov) was placed on the ground and the UAV was held above the
panel always keeping the same position—the Sun was behind the UAV, so no reflection
and shadow were affecting the panel, as this was recommended by the manufacturer. The
same rules were applied to the other cameras (e.g., [70]).
GPS data from DJI Phantom 4 and Parrot Sequoia camera were used to reference the
resulting orthomosaics into the coordinate system (WGS 84/UTM 33N). The imaging data
(RGB and multispectral bands) were processed in Agisoft Metashape software (Agisoft
LLC [71]), which allowed for orthomosaics (RGB and multispectral) and digital surface
models (DSMs) to be created using the structure from motion method [72]. The multispec-
tral data calibration was automatically made by the Agisoft Metashape software, which
detected the images of the calibrated panel by the QR code in it as described in [73].
2.4. Tree Height, Crown and Top Detection

Tree height detection was based on the Canopy Height Model (CHM) described as
the difference between treetop elevation and the underlying ground-level elevation [74].
In this study, the 5th Digital Elevation Model (DEM) generation (DMR 5G; [61,75]) was
used as a source of ground-level information, while treetop information was obtained
from the digital surface model (DSM) that was obtained from the Parrot Sequoia camera
multispectral images (Seq DSM):
CHM = Seq DSM − DMR 5G (1)
First, the DMR 5G data (original spatial resolution 0.5 m) were resampled to a Seq
DSM spatial resolution of 16–25 cm, and Seq DSM was calibrated by DMR 5G elevation to
obtain proper results. The calibration was performed at each site by extracting the average
heights from the Seq DSM of a clearing or forest path, then it was possible to compare this
height value with the average site altitude derived from the DMR 5G. The elevation of the
Seq DSM was then corrected by this difference.
The following procedure was employed to derive tree tops. Focal statistics was
employed to the CHM raster to get rid of noisy pixels and a new dataset was created
(CHMfoc). The best results were achieved when using the maximum value and a 5 × 5-
pixel window, according to the empirical tests carried out [76]. A window with a smaller
size caused frequent double peak detection. On the other hand, a bigger window did not
identify a sufficient number of tree peaks. The tree tops were detected by identifying the
Remote Sens. 2021, 13, 705 7 of 2
Remote Sens. 2021, 13, 705 5x5-pixel window, according to the empirical tests carried out [76]. A window 7 of 26 with a
smaller size caused frequent double peak detection. On the other hand, a bigger window
did not identify a sufficient number of tree peaks. The tree tops were detected by
identifying
local maximathe local
in the maxima
CHMfoc in the
raster. The CHMfoc raster.
representative Thefor
pixels representative pixels for the
the highest CHMfoc
highest CHMfoc values were then identified in the original CHM raster,
values were then identified in the original CHM raster, as it was important to keep the as it wa
important
original topositions
pixel keep theandoriginal
valuespixel positions
and the treetop and
pointvalues andderived
layer was the treetop point layer
[49]. Figure 3 wa
derived
shows the[49].
tree Figure 3 shows workflow.
height detection the tree height detection workflow.
Figure3.3.Workflow
Figure Workflowused in this
used study
in this to detect
study tree height,
to detect crown crown
tree height, and top.and top.
In order to delineate the tree crowns, the workflow described by Jaakkola et al. [50]
In order to delineate the tree crowns, the workflow described by Jaakkola et al. [50
was followed, which used watershed analysis to identify tree crown borders using the
was followed,
CHM which
derived from used watershed
UAV-based analysis
laser scanning. First,to identify
inverse CHMtree crownwas
(iCHM) borders using the
created,
where trees were visualized as depressions and treetops represented the lowest points in created
CHM derived from UAV-based laser scanning. First, inverse CHM (iCHM) was
where
the trees
digital were visualized
elevation model: as depressions and treetops represented the lowest points in
the digital elevation model: iCHM = CHM * (−1) (2)
The iCHM, together with the tree peak iCHM layer
= CHMrepresenting
* (−1) pour points, were used to (2
compute watershed regions. The watershed analysis showed good results, even in the
The
case of iCHM,
splitting together
two withnearby
or multiple the tree peak layer
standing representing
trees (Figure 4a). Treepour points,
borders were used to
in places
compute watershed regions. The watershed analysis showed good results, 4b),
with no connection to another tree crown were defined using a height mask (Figure even in the
more specifically
case of splitting excluding such areas
two or multiple where
nearby the CHM
standing altitude
trees was4a).
(Figure at least
Tree3borders
m lowerin place
than
withthe
noactual lowest detected
connection tree.tree
to another Thiscrown
allowed for the
were space under
defined using the tree crowns
a height maskto(Figure
be 4b)
removed. Consequently, to only visualise sunlit crown parts convenient for the following
more specifically excluding such areas where the CHM altitude was at least 3 m lowe
multispectral analyses, the shadow mask was derived by thresholding the Red band of
than the actual lowest detected tree. This allowed for the space under the tree crowns to
the multispectral mosaic (Seq Mosaic) and selecting values lower than 0.04 [77]. In the
be removed.
next Consequently,
step, the height to only
mask was merged withvisualise
the shadow sunlit
maskcrown
(Figureparts convenient
4c). The final tree for the
following
crown multispectral
boundaries analyses,
were obtained by the shadow
applying themask
mergedwas derived
mask to theby thresholding
watershed layer the Red
band of the multispectral mosaic (Seq Mosaic) and
(Figure 4d), as a result the individual tree crowns were extracted.selecting values lower than 0.04 [77]
In the next step, the height mask was merged with the shadow mask (Figure 4c). The
final tree crown boundaries were obtained by applying the merged mask to the
watershed layer (Figure 4d), as a result the individual tree crowns were extracted.
RemoteSens.
Remote Sens.2021, 13, 705
2021, 13, 705 8 of 26 8 of 2
Figure4.4.An
Figure An example
example of tree
of tree crown
crown detection
detection (tree peaks
(tree peaks visualized
visualized by greenby green
dots): (a) dots): (a) watershed
watersheds
surrounding
surrounding treetops
treetops (white
(white line)line)
splitsplit nearby
nearby trees—the
trees—the bordersborders are highlighted
are highlighted in red;
in red; DSM on DSM o
thebackground;
the background; (b)(b)
thethe height
height mask mask (yellow
(yellow line)one
line) cuts cuts one
side of side of the
the tree treein
crowns crowns
a placein a place
with a with
highaltitude
high altitude difference
difference (red(red highlight);
highlight); (c) the(c) the shadow
shadow mask line)
mask (orange (orange line)
reduces thereduces theofdark par
dark parts
of tree
tree crowns
crowns and completes
and completes the tree
the tree crown crown
borders; borders; multispectral
multispectral imagery from the imagery from the Parro
Parrot Sequoia
sensor on the background (false-colour composition: Green, Red, Red edge); and, (d) final resultand,
Sequoia sensor on the background (false-colour composition: Green, Red, Red edge); of (d) fina
result of tree crown
tree crown detection. detection.
2.5. Multispectral Data Processing

2.5. Multispectral Data Processing
The Parrot Sequoia camera captures four-band images (Green, Red, Red edge, and Near
The Parrot Sequoia camera captures four-band images (Green, Red, Red edge, an
infrared), which were specifically designed for vegetation analysis (Table 3). These bands allow
Near infrared),
computing, whichthe
for instance, were specifically
Normalized designed
Difference for vegetation
Vegetation analysis
Index (NDVI) [78],(Table
which3). Thes
isbands allowvegetation
a universal computing,index, for
butinstance,
also, morethe Normalized
importantly, Difference
they allow Vegetation Inde
the modification
of(NDVI) [78], using
NDVI when whichtheisRed
a universal
edge bandvegetation
instead of the index, but also,
Red band. moreNormalized
Red Edge importantly, the
allow theVegetation
Difference modificationIndexof(NDVI
NDVIred when using
edge ) has theused
been Redtoedge band
assess insteadstress
vegetation of the Red band
[79],
as the Edge
Red stress Normalized
directly affectsDifference
the wavelength positionIndex
Vegetation of the red-edge
(NDVIredinflexion
edge) has point
beenon the to asses
used
electromagnetic
vegetation stress spectrum
[79], [80,81].
as theThe NDVIdirectly
stress red edge index is defined
affects as:
the wavelength position of th
red-edge inflexion point on the electromagnetic
NIR − RE spectrum [80,81]. The NDVIred edge inde
is defined as: (3)
NIR + RE
where NIR is the Near infrared band and RE is the Red edge band of the Parrot Sequoia camera. (3
This study tested whether a linear relationship can be established between these
where
two NIR vegetation
popular is the Near infrared
indices (VIs)band andground
and the RE is the Red
truth edge
(e.g., band of the
biochemically Parrot Sequoi
assessed
camera.
needle photosynthetic pigment contents). The individual tree crown pixels were delineated
in the This study tested
multispectral imageswhether
while usinga linear relationshipapproach
the iCHM-based can be (described
establishedin between
detail thes
in Section 2.4). Moreover, the UAV-based imaging data provided at exceptionally
two popular vegetation indices (VIs) and the ground truth (e.g., biochemically assesse high
spatial
needleresolution provided an
photosynthetic opportunity
pigment to test whether
contents). the crown geometry
The individual tree crown and light
pixels wer
conditions affect the crown VI values and the estimation of the photosynthetic pigments
delineated in the multispectral images while using the iCHM-based approach (describe
(the validity of the linear models), respectively. Therefore, the intention was to delineate
in detail in Section 2.4). Moreover, the UAV-based imaging data provided a
crown parts that receive higher and lower illumination. First, Principal Component Anal-
exceptionally
ysis highemployed
(PCA) [82] was spatial resolution provided
to reduce the an opportunity
four-band dimensionality tointo
testthree
whether
bands the crow
(PC1–PC3) and display them as an RGB composition. PC1–PC3 were then visually assessed of th
geometry and light conditions affect the crown VI values and the estimation
photosynthetic
to check whether it pigments
is possible(the validitytheofdetected
to classify the linear
treemodels), respectively.
crowns according Therefore, th
to different
intention was
illumination to delineate
conditions. crown parts
For simplicity, thatthree
these receive higher
Principal and lower (PC1–PC3),
Components illumination. Firs
which covered
Principal the majorityAnalysis
Component of the information
(PCA) [82] variability, were selected
was employed to as the input
reduce thefor
four-ban
the subsequent unsupervised classification employing the Iterative Self-Organizing
dimensionality into three bands (PC1–PC3) and display them as an RGB composition Data
Analysis
PC1–PC3 Technique
were then(ISODATA;
visually [83]). A preliminary
assessed to checkISODATA
whether classification
it is possiblewas initially
to classify th
set to yield a maximum of four classes; these classes were then recoded into two classes
detected tree crowns according to different illumination conditions. For simplicity, thes
three Principal Components (PC1–PC3), which covered the majority of the informatio
variability, were selected as the input for the subsequent unsupervised classificatio
employing the Iterative Self-Organizing Data Analysis Technique (ISODATA; [83]). A
preliminary ISODATA classification was initially set to yield a maximum of four classe
Remote Sens. 2021, 13, 705 9 of
Remote Sens. 2021, 13, 705 9 of 26
using the RGB ortho-photomosaics, which were provided at the highest spati
based on thefor
resolution visual inspection
all image when using the RGB ortho-photomosaics, which were
products.
provided at the highest spatial resolution for all image products.
It was then possible to define three scenarios for further analysis of th
It was then possible to define three scenarios for further analysis of the multispectral
data (Figure 5): data (Figure 5):
multispectral
• • Scenario
Scenario 1: all
1: all of the
of the pixels
pixels representing
representing the whole
the whole tree have
tree crown crownbeenhave been average
averaged
and
and used
used forfor further
further statistical
statistical analysis.
analysis.
• • Scenario
Scenario 2: pixels representinghigher-illumination
2: pixels representing the the higher-illumination
top parttop partcrown
of the of thehave
crown hav
been averaged
been averaged andandusedused
for further statistical
for further analysis.
statistical analysis.
• • Scenario 3: pixels
Scenario thatthat
3: pixels represent the lower-illumination
represent part of the
the lower-illumination partcrown
of thehave beenhave bee
crown
averaged and used for further statistical analysis.
averaged and used for further statistical analysis.
Figure
Figure 5. 5. Parrot
Parrot Sequoia
Sequoia multispectral
multispectral data processing
data processing workflow
workflow chart. chart.
For
Foreach scenario,
each a linear
scenario, model
a linear between
model the derived
between VI values
the derived VI and corresponding
values and correspondin
ground truth (needle photosynthetic pigment contents defined in the laboratory) was
ground truth (needle photosynthetic pigment contents defined in the laboratory) wa
individually constructed. Different needle age generations contribute to the final image
individually constructed. Different needle age generations contribute to the final imag
pixel reflectance; however, it is still not clear what the optimal ground truth age proportion
is,pixel reflectance;
as already however,
mentioned. it is still not
The recommendation clear
given what
by [60] wasthe optimal
followed ground
to include nottruth ag
proportion is, as already mentioned. The recommendation given by [60]
only current (first year) but also older needles as a ground truth and four ground truthwas followed
include not
validation only current
age groups (first
were thus year) but also older needles as a ground truth and fou
created:
• ground truth included
all needles validation age groups were thus created:
• • first
allyear needles
needles included
included
• • second
first year needlesincluded
year needles included
• second year needles included
• mixed sample of fourth year and older needles (hereinafter referred to as four
year for simplicity)
To evaluate the linear models, the coefficient of determination (R2) was used:
Remote Sens. 2021, 13, 705 10 of 26
• mixed sample of fourth year and older needles (hereinafter referred to as fourth year
for simplicity)
To evaluate the linear models, the coefficient of determination (R2 ) was used:
SSres
R2 = 1 − (4)
SStot
where SSres represents the residual sum of squares and SStot represents total sum of
squares [84].
3. Results
3.1. Photosynthetic Pigments
Laboratory analyses of needles that were collected as a ground truth showed typical
chlorophyll and carotenoid values (Table 4) for non-stressed mature Norway spruce trees
in a similar region and at a similar altitude [60,85]. This was despite the fact that forest in
the PLB catchment showed evidence of suppressed growth due to the nutritional stress that
is caused by the extreme chemistry of the underlying bedrock [86]. This further indicated
that photosynthetic pigment content alone provided a somewhat limited indication of
stressed trees. At both of the study sites, almost all trees exhibited the usual accumulation
of chlorophyll and carotenoids in older needles in comparison with the first year needles
(Table 4) [22,85,87–90].
Table 4. Mean contents of photosynthetic pigments based on in-situ needle samples for both catchments.
Catchment All Needles 1st Year Needles 2nd Year Needles 4+ Years Needles
Total Chlorophyll Lysina (LYS) 52.580 35.756 56.055 65.399
(µg·cm−2 ) Pluhův Bor (PLB) 44.425 31.148 43.895 57.172
Chlorophyll a Lysina (LYS) 37.939 26.124 40.679 46.637
Chlorophyll b Lysina (LYS) 14.641 9.631 15.373 18.761
Carotenoids Lysina (LYS) 6.724 4.400 7.018 8.602
3.2. UAV Photogrammetric Products

The UAV data were acquired over three stands in the Lysina catchment (LYS 1K, LYS
2K, and LYS 4K) and three stands in the Pluhův Bor catchment (PLB 2K, PLB 3K, and PLB
5K). At each stand, the DSM, RGB, and multispectral orthomosaics were created (Figure 6).
Table 5 shows the resulting spatial resolution and photogrammetric model errors. The total
errors of the photogrammetric products oscillated around 1 m. The calculated vertical error
of DSMs varied between 0.51 and 0.78 m, and the obtained accuracy was comparable with
the results that were published for an DJI Phantom 4 UAV [91]. The RGB data from DJI
Phantom 4 were then mainly used for a visual control, while the calibrated multispectral
data were used for further statistical assessment.
Remote Sens. 2021, 13, 705 11 of 26
ns. 2021, 13, 705 11 of 28
Figure(LYS
Figure 6. Example 6. Example (LYSsurvey
2K) of the UAV 2K) of the UAV
products survey
used in products
this study: usedorthomosaic
(a) the RGB in this study: (a)from
obtained the the
RGBDJI
orthomosaic obtained from the DJI camera; (b) the multispectral mosaic derived from the Parrot
camera; (b) the multispectral mosaic derived from the Parrot Sequoia sensor (false-colour composition: Green, Red, Red
edge); and, Sequoia sensor
(c) the digital (false-colour
surface composition:
models (DSM) derived fromGreen, Red,
the Parrot Red multispectral
Sequoia edge); and, data.
(c) the digital surface
models (DSM) derived from the Parrot Sequoia multispectral data.
Table 5. The resulting spatial resolution and photogrammetric model errors of multispectral and RGB data.
Table 5. The resulting spatial resolution and photogrammetric model errors of multispectral and RGB data.
Parrot Sequoia-Multispectral Camera DJI Camera-RGB
Parrot Sequoia - Multispectral Camera DJI Camera - RGB
Orthomosaic
DSM (cm/px)
Orthomosaic Total Error Orthomosaic Total Error
Test
DSM Site(cm/px) (cm/px) Vertical Error (m) Error
Total Orthomosaic Total Error(m)
t Site (Seq DSM)(cm/px) (Seq Mosaic)
Vertical Error (m) (m) (cm/px)
(Seq DSM) (m) (cm/px) (m)
LYS 1K (Seq
23.95 Mosaic) 5.99 0.78 1.02 2.26 0.77
S 1K 23.95
LYS 2K 16.41 5.99 4.09 0.78 0.55 1.02 0.67 2.26 1.12 0.77 1.14
S 2K LYS 4K
16.41 19.64 4.09 4.91 0.55 0.65 0.67 0.94 1.12 2.17 1.14 0.79
S 4K PLB 2K
19.64 23.30 4.91 5.82 0.65 0.51 0.94 0.82 2.17 2.05 0.79 0.71
PLB 3K 24.47 6.11 0.53 0.74 2.36 1.27
B 2K 23.30
PLB 5K 23.00
5.82 5.74
0.51 0.61
0.82 0.72
2.05 1.87
0.71 0.83
B 3K 24.47 6.11 0.53 0.74 2.36 1.27
B 5K 23.00 5.74 0.61 0.72 1.87 0.83
3.3. UAV Tree Height, Crown and Top Detection
3.3. UAV Tree Height,
The Crown and Top Detection
tree characteristics—height, crown, and top—derived from the CHM raster
The tree(Figure 7) were compared with the in-situ measurements that were collected in 2015 [62].
characteristics—height, crown, and top—derived from the CHM raster
Table 6 summarizes the evaluation statistics. One stand—LYS 1K—showed very high
(Figure 7) were compared with the in-situ measurements that were collected in 2015 [62].
error; for the other stands, the success rate for the tree top identification varied between
Table 6 summarizes
72 and 87%.the When
evaluation statistics.
checking One
the in-situ stand—LYS
data for LYS 1K,1K—showed
it was found thatverythis
high
stand is
error; for the represented
other stands, the success rate for the tree top identification varied between
by significantly younger trees with the highest trees measured around 12 m.
72 and 87 %. At When checkingstand
this particular the in-situ
the treesdata
werefor LYS
very 1K,and
dense it was found
rather short,that this stand
therefore is
tree detection
represented by significantly
could younger
be problematic. trees
As this with
stand the
also highestvery
obtained trees measured
high errors in around
all other 12 m.
statistics, it
was defined as an outlier and excluded from further statistics.
At this particular stand the trees were very dense and rather short, therefore tree In the case of tree crown
detection could estimation, the comparable
be problematic. As this success
standrate achieved
also varied
obtained verybetween 76–94%in(Table
high errors 6). In two
all other
cases, PLB 3K and PLB 5K, the tree crown detection results showed a higher number of es-
statistics, it was defined as an outlier and excluded from further statistics. In the case of
timated numbers than those that were measured in the field. This could be possibly caused
tree crown estimation, the comparable success rate achieved varied between 76–94%
by false crown splitting when using the tree shadow mask in the detection algorithm.
(Table 6). In two cases, PLB 3K and PLB 5K, the tree crown detection results showed a
higher number of estimated numbers than those that were measured in the field. This
could be possibly caused by false crown splitting when using the tree shadow mask in
the detection algorithm.
Remote Sens. 2021, 13, 705 12 of 26
Remote Sens. 2021, 13, 705 12 of 28
Figure 7.
Figure 7. Example
Example of
of result
resultdata
datafrom
fromtree
treepeak,
peak,height
heightand
and crown
crown detection
detection process:
process: (a)(a)
LYSLYS
1K 1K
testtest location;
location; (b) LYS
(b) LYS 2K
2K test
test location. Green dots represent estimated tree peaks, yellow polygons delineate detected tree crowns and red-line
location. Green dots represent estimated tree peaks, yellow polygons delineate detected tree crowns and red-line polygons
polygons show the three trees at each stand which were sampled for photosynthetic pigment contents. The background
show the three trees at each stand which were sampled for photosynthetic pigment contents. The background image is the
image is the Multispectral Parrot Sequoia mosaic (false-colour composition: Green, Red, Red edge).
Multispectral Parrot Sequoia mosaic (false-colour composition: Green, Red, Red edge).
Table 6. Comparison between the number of detected tree peaks and crowns (Canopy Height Model (CHM)-estimated)
Table 6. Comparison
and in-situ measuredbetween
data. the number of detected tree peaks and crowns (Canopy Height Model (CHM)-estimated) and
in-situ measured data.
Success Rate of
Success Rate of
No. of Trees Measured No. of Detected Tree No. of Detected Tree Detected
Success Tree
RateCrowns
of
Detected
Success Tree
Rate of Tops (%) Crowns Based on UAV (%) (Compared to the
Test Site In-Situ
No. of Trees(Above Set Peaks Based on UAV No. of Detected Detected Tree
(Compared to the Trees
MeasuredHeight
In-SituMask) No. of Detected
Data Detected Tree Tops Data
Tree Crowns Trees Measured
Crowns (%)
Test Site Tree Peaks Based Measured
(%) (Compared toIn-Situ)
the
(Above Set Height Based on UAV In-Situ)
(Compared to the
LYS 1K on UAV Data45 Trees Measured
Mask) 177 25.42 Data 61 34.46
Trees Measured
LYS 2K 67 49 In-Situ)73.13 57 85.07
In-Situ)
LYS 4K 25 18 72.00 19 76.00
LYS PLB
1K 2K 177 34 45 25 25.42 73.53 61 31 34.46
91.18
LYS PLB
2K 3K 67 15 49 13 73.13 86.66 57 17 85.07
86.67
LYS PLB
4K 5K 25 32 18 26 72.00 81.25 19 34 76.00
93.75
PLB 2K 34 25 73.53 31 91.18
PLB 3K 15 The box 13 plots showing 86.66 the distributions of 17 tree height values 86.67 for all
PLB 5K 32 stands—CHM-derived
26 and in-situ—are
81.25 displayed in 34 Figure 8. Furthermore, 93.75 either all
trees were validated (LYS 4K—18 trees and PLB 3K—13 trees) or a comparable yet
representative number—in this case, the 20 tallest trees—were selected to validate the
The box plots showing the distributions of tree height values for all stands—CHM-
bigger
derivedstands (LYS 2K, PLB
and in-situ—are 5K, and
displayed in PLB
Figure2K). The averagedeither
8. Furthermore, CHM-estimated
all trees weretree height
validated
values were compared with the in-situ data. It can be concluded
(LYS 4K—18 trees and PLB 3K—13 trees) or a comparable yet representative number—in that the tree height
values that
this case, thewere derived
20 tallest for these selected
trees—were trees from CHM were
to validate alwaysstands
the bigger higher(LYS
than2K,
thePLB
in-situ
5K,
measurements, the differences ranged between 1.54–2.36 m (Table
and PLB 2K). The averaged CHM-estimated tree height values were compared with the 7). These differences
are most
in-situ probably
data. It canabe combination
concluded of thatthethe
vertical error of
tree height the Seq
values thatDSM
were(0.51–0.78
derived for m; Table
these
5) and tree growth between the years 2015 and 2018. Based on a relationship
trees from CHM were always higher than the in-situ measurements, the differences ranged between the
tree diameter at breast height (DBH) and height [62], the 2015–2018
between 1.54–2.36 m (Table 7). These differences are most probably a combination of the average height
increment
vertical error was estimated
of the Seq DSMto(0.51–0.78
be 1.1+-0.5 m at 5)Lysina
m; Table and
and tree 0.4+-0.2
growth m atthe
between Pluhův Bor.
years 2015
However, the correlation between the estimated heights of the
and 2018. Based on a relationship between the tree diameter at breast height (DBH) and 20 tallest trees using
CHM
heightand [62],in-situ measuredaverage
the 2015–2018 data was high increment
height (Figure 9), was
as the coefficients
estimated of 1.1
to be determination
± 0.5 m at
(R 2) obtained were higher than 0.90 in four out of five cases.
Lysina and 0.4 ± 0.2 m at Pluhův Bor. However, the correlation between the estimated
heights of the 20 tallest trees using CHM and in-situ measured data was high (Figure 9),
as the coefficients of determination (R2 ) obtained were higher than 0.90 in four out of
five cases.
Remote Sens. 2021, 13, 705 13 of 2
Remote
Remote Sens.2021,
Sens. 13, 705
2021,13, 705 13 of 26 13 of 2
Figure 8. Tree height distribution visualized by box plots. Blue boxes represent trees detected b
the
Figureautomatic
Figure 8.8.Tree
Tree algorithm
height
height from
distribution
distribution UAV-based
visualized
visualized CHM
by box by data
boxBlue
plots. plots. (2018)
boxesBlue and trees
boxes
represent orange
represent boxes
detectedtrees display
detectedth
by the b
distribution
the automatic
automatic of the
algorithm tree heights
algorithm
from frommeasured
UAV-based UAV-based
CHM datain-situ
(2018)inand
CHM 2015.
data
orange(2018) and orange
boxes display boxes display
the distribution of th
distribution
the tree heightsofmeasured
the tree heights
in-situ inmeasured
2015. in-situ in 2015.
Table 7. Comparison of average tree heights of the 20 tallest trees for each stand: CHM-estimated
tree
Table
Table heights and theofsitu
7.7.Comparison
Comparison measured
of average
average data.
treetree heights
heights of theof20the 20 tallest
tallest trees
trees for forstand:
each each CHM-estimated
stand: CHM-estimated
treeheights
tree heights and
and thethe
situsitu measured
measured data.data.
Average Tree Top Average Tree Top Difference of
Test Site Average
Height Tree
(m)
Average
Top Average
- Top Height
Tree (m)Tree
Average TreeTop
BasedTopon Difference
Average of
Tree
Difference of
Test Site Height (m) Based on Average Tree
Test Site Height
In-Situ
Height (m) - DataHeight
Data
(m)-In-Situ (m) Based
the CHM Data on Average
Heights Tree
(m)
the CHM Data Heights (m)
LYS 2K In-Situ
13.33Data the CHM
14.87 Data Heights
1.54 (m)
LYS 2K 13.33 14.87 1.54
LYS 2K
LYS4K
4K 13.33
24.3424.34 14.87
26.62
26.62 1.54
2.28
2.28
LYS 2K
PLB
PLB 4K
2K 24.34
27.0827.08 26.62
29.44
29.44 2.28
2.36
2.36
PLB 3K 22.86 24.63 1.77
PLB 2K
3K 27.08
22.86 29.44
24.63 2.36
1.77
PLB 5K 23.95 25.70 1.75
PLB 5K3K 22.86
23.95 24.63
25.70 1.77
1.75
PLB 5K 23.95 25.70 1.75
Figure9.9.Linear
Figure Linear regression
regression andand coefficients
coefficients of determination (R2 ): the(R
of determination 2): the CHM-estimated heights of
CHM-estimated heights of the
the
20 20 highest
highest
Figure 9. trees trees
at each
Linear at each stand
stand compared
regression compared to the
to the in-situ
and coefficients in-situ
data.
of determinationdata.(R ): the CHM-estimated heights of
2
the 20 highest trees at each stand compared to the in-situ data.

3.4. Tree Crown Illumination Classes
3.4. Tree
TheCrown Illumination
first three Classes
Principal Components (PC1–PC3) covered more than 99% of th
crown spectral
The variability
first three (Table
Principal 8), and they(PC1–PC3)
Components were usedcovered
as the input
moreforthan
the subsequen
99% of th
Remote Sens. 2021, 13, 705 14 of 26
3.4. Tree Crown Illumination Classes

The first three Principal Components (PC1–PC3) covered more than 99% of the crown
spectral variability (Table 8), and they were used as the input for the subsequent unsuper-
vised ISODATA classification. The classification results were visually validated using the
ortho-photomosaics (RGB and multispectral), and the best results were achieved using the
following set up (maximum iterations 30, maximum class standard deviation 1, maximum
class distance 5, and change threshold 5%). It was concluded that the final classification
enabled a differentiation between the higher (Scenario 2) and lower-illuminated (Scenario 3)
parts of a crown. Figure 10a,b give an example of tree No. 22 (PLB 5K stand), showing the
distribution of the two final illumination classes within the tree crown and Figure 10c shows
the mean reflectance derived for the two illumination scenarios per watershed. Clearly,
Scenario 2 contains pixels that receive higher-illumination (overall higher reflectance, sig-
nificantly higher mean, minimum, and maximum values in the Red Edge-NIR region) and
Scenario 3 includes pixels that receive the lower illumination (overall lower reflectance,
significantly lower mean, minimum, and maximum values in the Red Edge-NIR region).
Zhang et al. [92] described the same trend at the leaf level.
Table 8. Principal Component Analysis (PCA) statistics obtained for tree crowns analysed at the two
sites: Lysina (LYS) and Pluhův Bor (PLB) catchments.
LYS
PC Component Eigenvalue Variance Cumulative Variance
1 0.01602 98.6427 98.6427
2 0.00015 0.9405 99.5832
3 0.00006 0.3733 99.9566
4 0.00001 0.0434 100
PLB
PC Component Eigenvalue Variance Cumulative variance
1 0.021 99.0774 99.0774
2 0.00014 0.6383 99.7157
3 0.00005 0.2335 99.9493
4 0.00001 0.0507 100
3.5. Relationship between Selected Vegetation Indexes and the Ground Truth
Both of the tested catchment areas were characterized by comparable NDVIred edge
values; however, a bigger difference in the NDVI values could be seen between the LYS
and PLB catchments (Table 9). In the methodological chapter (Section 2.5), three differ-
ent scenarios were defined to assess the relationship between the spectral indices and
photosynthetic pigments: Scenario 1: the whole crown (all pixels); Scenario 2: the top
and the higher-illuminated part of the tree crown; and, Scenario 3: the lower-illuminated
part of the tree crown. When comparing the results that were obtained for Scenarios 1–3
(Table 9, Figure 11), the NDVI index showed an almost negligible differences among all
three Scenarios, while the NDVIred edge index showed minimally larger differences among
the defined scenarios.
The linear models were built between the VIs (NDVI and NDVIred edge ) and the
ground truth (in-situ needle chlorophyll and carotenoid contents). The LYS 1K stand was
identified as an outlier, as mentioned in Section 3.3. Besides the significant differences in
tree age (Table 2), this stand was also characterized by a much higher tree density (field
observations—LAI at LYS 1K = 4.47 in comparison to LYS 2K = 3.58 and LYS 4K = 3.70).
High tree density indicated high VIs values; however, it did not correspond with rather low
values of chlorophyll content. Therefore, LYS 1K was excluded from the further statistics.
As a result, there were nine trees in total from the PLB catchment and six trees from the
LYS catchment that were used for this statistical assessment. Three different scenarios
the tree crown and Figure 10c shows the mean reflectance derived for the two
illumination scenarios per watershed. Clearly, Scenario 2 contains pixels that receive
higher-illumination (overall higher reflectance, significantly higher mean, minimum,
and maximum values in the Red Edge-NIR region) and Scenario 3 includes pixels that
receive the lower illumination (overall lower reflectance, significantly lower mean,
minimum, and maximum values in the Red Edge-NIR region). Zhang et al. [92]
Remote Sens. 2021, 13, 705 15 of 26
described the same trend at the leaf level.
Table 8. Principal Component Analysis (PCA) statistics obtained for tree crowns analysed at the
two sites: Lysina (LYS) and Pluhův Bor (PLB) catchments.
were used (Figures 12 and 13), LYS for which the VIs were correlated with the ground truth
(laboratory
PC Componentanalysis of Chlorophyll a and
Eigenvalue b and carotenoids)
Variance using
Cumulative four different needle age
Variance
groups1 too—all needles 0.01602 98.6427
together included, 98.6427
first year needles included, second year needles
2 0.00015 0.9405 99.5832
included,
3
and four year old needles included,
0.00006 0.3733
as already 99.9566
described (Tables 10 and 11). A
similar4 pattern was identified,
0.00001 for both Vis.
0.0434 It can be concluded
100 that the results differ
significantly depending on what needle age group was used as ground truth. The worst
PLB
PC Component
results were obtainedEigenvalue
when the first year Variance
needle group Cumulative variance
was used (no correlation), followed
1 0.021 99.0774 99.0774
by the2 four year old needle
0.00014 group and all age
0.6383 needle group used as ground truth (very
99.7157
weak correlation).
3 On 0.00005
the other hand, for0.2335
both VIs, the best99.9493
results were obtained when the
second4 year needle group
0.00001was used as ground
0.0507 truth. 100
Remote Sens. 2021, 13, 705 15 of 28
representing Scenario 2 (the top higher-illumination part of the crown) and Scenario 3 (the
lower-illumination part of the crown); and, (c) mean reflectance per two illumination scenarios.
Vertical bars represent minimum and maximum values for each band and scenario.
3.5. Relationship between Selected Vegetation Indexes and the Ground Truth
Both of the tested catchment areas were characterized by comparable NDVIred edge
values; however, a bigger difference in the NDVI values could be seen between the LYS
and PLB catchments (Table 9). In the methodological chapter (Section 2.5), three
different scenarios were defined to assess the relationship between the spectral indices
and photosynthetic pigments: Scenario 1: the whole crown (all pixels); Scenario 2: the
top
Figure
Figure
and10.the
10. Scheme
higher-illuminated
Scheme
of Tree of
No.Tree
22 atNo.
the 22
PLBat5K
the
part of the
PLBshowing
stand 5K stand
treeSequoia
showing
(a) Parrot
crown;
(a) Parrot
and, Scenario 3: the
Sequoia multispectral data
multispectral
lower-illuminated
data in band combination part
4-2-1 of
(NIR, the
Red, tree
Green) crown. When
corresponding to comparing
Scenario 1; (b) two the
classes results
in band combination 4-2-1 (NIR, Red, Green) corresponding to Scenario 1; (b) two classes representing
that were
obtained for Scenarios 1–3 (Table 9, Figure 11), the NDVI index showed
Scenario 2 (the top higher-illumination part of the crown) and Scenario 3 (the lower-illumination
an almost
negligible differences among all three Scenarios, while the NDVI red edge index showed
part of the crown); and, (c) mean reflectance per two illumination scenarios. Vertical bars represent
minimally larger differences among the defined scenarios.
minimum and maximum values for each band and scenario.
Table 9.9.
Table Mean
Meanvalues/Standard
values/Standard deviations of vegetation
deviations indices
of vegetation based
indices on on
based Sequoia optical
Sequoia data
optical perper
data
evaluated
evaluatedScenarios forfor
Scenarios both catchments.
both catchments.
Catchment
Catchment ScenarioScenario
1 1 Scenario 2
Scenario 2 Scenario 3 3
Scenario
Lysina (LYS) 0.831/0.022 0.832/0.022 0.829/0.023
NDVI Lysina (LYS) 0.831/0.022 0.832/0.022 0.829/0.023
NDVI Pluhův Bor (PLB)
Pluhův Bor (PLB)0.782/0.056
0.782/0.0560.772/0.055
0.772/0.055 0.770/0.059
0.770/0.059
Lysina (LYS)
Lysina (LYS) 0.138/0.022
0.138/0.0220.130/0.019
0.130/0.019 0.149/0.025
0.149/0.025
NDVI
NDVI red edge
red edge
Pluhův Pluh
Bor ův Bor (PLB)0.140/0.019
(PLB) 0.140/0.0190.121/0.018
0.121/0.018 0.146/0.020
0.146/0.020
Figure 11. Normalized Difference Vegetation Index (NDVI), NDVIred edge : mean value and standard
Figure 11. Normalized Difference Vegetation Index (NDVI), NDVIred edge: mean value and standard
deviation
deviation forfor two
two illumination
illumination Scenarios
Scenarios 2 and
2 and 3 (summarized
3 (summarized forfor each
each catchment
catchment area).
area).
The linear models were built between the VIs (NDVI and NDVIred edge) and the
ground truth (in-situ needle chlorophyll and carotenoid contents). The LYS 1K stand
was identified as an outlier, as mentioned in Section 3.3. Besides the significant
differences in tree age (Table 2), this stand was also characterized by a much higher tree
Remote Sens. 2021, 13, 705 18 of 28
Remote Sens. 2021, 13, 705 16 of 26
Figure 12. Linear regression between NDVI index and second year needles photosynthetic pigment content for Scenarios 1–3.
Figure 12. Linear regression between NDVI index and second year needles photosynthetic pigment content for Scenarios
1–3.
Remote Sens. 2021, 13, 705 19 of 28
Remote Sens. 2021, 13, 705 17 of 26
Figure 13. Linear regression between NDVIred edge index and second year needles’ photosynthetic pigment content for
Figure 13.
Scenarios 1–3.Linear regression between NDVIred edge index and second year needles’ photosynthetic pigment content for
Scenarios 1–3.
second year needle group was used as ground truth together with Scenario 3 for both
VIs. Using this setting for the NDVI index, the following R2 were obtained for
photosynthetic pigments (Table 10): total chlorophyll—R2 = 0.49, Chlorophyll a—R2 =
0.48, Chlorophyll b—R2 = 0.51, and Carotenoids—R2 = 0.50. Comparable results were
Remote Sens. 2021, 13, 705 then also obtained for the NDVIred edge index (Table 11): total chlorophyll—R2 = 0.46,
18 of 26
Chlorophyll a—R2 = 0.45, Chlorophyll b—R2 = 0.48, and Carotenoids—R2 = 0.52. The
NDVI index achieved a slightly higher R2 for Total chlorophyll, Chlorophyll a and b
contents, while NDVIred edge showed slightly better results for Carotenoids.
Table 10. Coefficients of determination (R2 ) for NDVI index and the four ground truth needle
Table 10. Coefficientsage groups.
of determination (R2) for NDVI index and the four ground truth needle age groups.
Ground Truth Age Ground Truth Age

Parameter Scenario 1
Parameter Scenario
Scenario 1 2 Scenario 2 Scenario 3
Scenario 3
Group Group
Total chlorophyll 0.33284
0.33284 0.3132 0.34764
0.34764
Chlorophyll a 0.31794
0.31794 0.30137 0.33012
0.33012
All needles All b
needles
Chlorophyll 0.36924
Chlorophyll b 0.34138
0.36924 0.34138 0.39117
0.39117
Carotenoids 0.31854
Carotenoids 0.2931
0.31854 0.2931 0.33969
0.33969
Total chlorophyll 0.03731 0.04066 0.03605
Chlorophyll a 0.03489 0.03854 0.03333
1st year needles Chlorophyll a 0.03489 0.03854 0.03333
Chlorophyll b needles
1st year 0.04373 0.04618 0.04336
Chlorophyll b 0.04373 0.04618 0.04336
Carotenoids 0.07389 0.0579 0.08686
Total chlorophyll Carotenoids
0.4801 0.07389
0.44667 0.0579 0.08686
0.49043
Chlorophyll A Total chlorophyll
0.47219 0.4801
0.44141 0.44667 0.49043
0.48051
2nd year needles Chlorophyll A 0.47219 0.44141 0.48051
Chlorophyll B needles
2nd year 0.49659 0.45638 0.51224
Carotenoids Chlorophyll B
0.48543 0.49659
0.44873 0.45638 0.51224
0.50073
Total chlorophyll Carotenoids
0.21125 0.48543
0.19238 0.44873 0.50073
0.23391
0.21125 0.19238 0.21753
0.23391
4th year needles
0.19768 0.18171 0.27382
0.21753
4th year needles
Carotenoids 0.18607
0.24389 0.21744 0.21334
0.27382
Carotenoids 0.18607 0.16277 0.21334
Legend: .
Legend: .
Table 11. Coefficients of determination

Table 11. (R
2) for NDVIred edge index, and the2four ground truth age groups.
Coefficients of determination (R ) for NDVIred edge index, and the four ground truth
Ground Truth Age age groups.
Parameter Scenario 1 Scenario 2 Scenario 3
Group
Ground Truth Age
Total chlorophyll Parameter
0.33424 Scenario 1
0.3367 Scenario 2 0.32031
Scenario 3
Group
Chlorophyll a 0.32371 0.32902 0.30735
All needles Total chlorophyll 0.33424 0.3367 0.32031
Chlorophyll b 0.35841 0.35316 0.35132
Chlorophyll a 0.32371 0.32902 0.30735
Carotenoids
All needles 0.37654 0.37009 0.36017
Total chlorophyll Chlorophyll
0.03079 b 0.35841
0.05179 0.35316 0.018750.35132
Chlorophyll a Carotenoids
0.03029 0.37654
0.05143 0.37009 0.01810.36017
1st year needles Total chlorophyll 0.03079 0.05179
Chlorophyll b 0.03195 0.05241 0.01810.01875
Carotenoids Chlorophyll a
0.07625 0.03029
0.09464 0.05143 0.073310.0181
Remote Sens. 2021, 13, 705 1st year needles 17 of 28
Chlorophyll b 0.03195 0.05241 0.0181
Carotenoids 0.07625 0.09464 0.07331
Chlorophyll a 0.44209 0.42862 0.45333
2ndChlorophyll
year needlesa 0.44209 0.42862 0.45333
2nd year needles Chlorophyll b 0.46151 0.43743 0.48268
Chlorophyll b 0.46151 0.43743 0.48268
Carotenoids 0.51865 0.50788 0.52091
Carotenoids 0.51865 0.50788 0.52091
Total chlorophyll Total chlorophyll
0.2499 0.2499 0.22732
0.22732 0.24152
0.24152
Chlorophyll a 0.23973 0.22026 0.22884
4thChlorophyll
year needlesa 0.23973 0.22026 0.22884
4th year needles Chlorophyll b 0.27252 0.24211 0.27104
Chlorophyll b 0.27252 0.24211 0.27104
Carotenoids Carotenoids
0.26164 0.26164 0.23661
0.23661 0.25304
0.25304
Legend: .
Legend: .
Assessing the different illumination crown conditions (Scenarios 1–3) in the case that the
second year needle group was selected as ground truth (Tables 10 and 11, Figures 12 and 13),
the worst results were surprisingly obtained for Scenario 2, where the top crown part
receiving the higher-illumination was assessed, followed by Scenario 1—the whole crown
Remote Sens. 2021, 13, 705 19 of 26
case; the best results were obtained for Scenario 3—where the lower illumination part
of the crown was assessed. For the NDVI index, the differences between Scenarios 1–3
are very small, slightly bigger differences could be found for the NDVIred edge index. It
can be concluded that the best results were achieved when the second year needle group
was used as ground truth together with Scenario 3 for both VIs. Using this setting for
the NDVI index, the following R2 were obtained for photosynthetic pigments (Table 10):
total chlorophyll—R2 = 0.49, Chlorophyll a—R2 = 0.48, Chlorophyll b—R2 = 0.51, and
Carotenoids—R2 = 0.50. Comparable results were then also obtained for the NDVIred edge
index (Table 11): total chlorophyll—R2 = 0.46, Chlorophyll a—R2 = 0.45, Chlorophyll
b—R2 = 0.48, and Carotenoids—R2 = 0.52. The NDVI index achieved a slightly higher R2
for Total chlorophyll, Chlorophyll a and b contents, while NDVIred edge showed slightly
better results for Carotenoids.
4. Discussion
Regarding the first part of this analysis—tree height, crown, and top detection—it
can be concluded that the obtained results satisfy the requirements of this study, which
was basically aimed at identifying individual tree crowns and masking the background
and shades. A relatively quick method was employed using Seq DSM and high resolution
DEM (DMR 5G), and reliable results were obtained achieving R2 = 0.90 and higher between
CHM-estimated heights and the in-situ tree measurements for most of the stands, with
the exception of PLB 2K (R2 = 0.7) (Figure 9). The accuracy of tree-height estimation was
comparable or even higher than in the case of Pinus ponderosa forest (R2 = 0.71; [36]), where
the same multispectral camera was used, although with a different type of UAV. The lower
accuracy that was obtained for PLB 2K can be explained by changes happening between
the time the in-situ data were collected (2015) and the UAV data acquisition (2018). In the
field. it was observed that tree cuts were common in this particular forest stand. It will be
interesting to test in the future, if this approach can be further employed or adjusted for
estimating the forest aboveground biomass in a similar way as airborne laser scanning has
been employed [62].
For both VIs a similar R2 was obtained between the index values and the laboratory
analysis of photosynthetic pigments, NDVI showed a slightly higher R2 for chlorophyll
content. while NDVIred edge had the highest R2 for carotenoid content. The Red edge is
commonly used for detecting vegetation stress [93] and, in this study, it was also a slightly
better index to estimate carotenoids—the vegetation stress indicators. The biggest influence
on linear models for both was the selection of the needle age group used as a ground truth.
As summarized in Tables 10 and 11, basically the only usable results were obtained when
using the second year needle pigment contents. The age-dependence of correlation strength
between vegetation indices and measured pigment contents were expectable. However, the
absence of a NDVI and NDVIred edge correlation to chlorophyll and carotenoids for the first
year needles was highly surprising. Such needles were already fully developed at the date
of sampling (August 4-5th), thus the immaturity of needles as a likely reason was excluded.
Moreover, first year needles are routinely and successfully used for taking ground truth for
a broad range of spectroscopic and remote sensing studies at various scales from leaf- to
stand level [89,93–97].
However, in some studies, second year needles, similarly as in this study, also proved
to be the best option for predicting needle traits [60,98], although the physiological and
optical causes have not been elucidated. The second year needles were also successfully
used as ground truth for pigment content estimation from multispectral UAV data in
mature Scots spine (Pinus sylvestris) [96]. The authors justified the selection of the second
year needles, literally as “avoiding non-representative outliers in current and mature nee-
dles” [96]. The absence of correlation between chlorophyll content in first year needles and
vegetation indices could be partly related to other interfering needle traits than chlorophyll
content itself. Although water has absorption features in near- and shortwave infrared
region, it sometimes shows an intercorrelation with chlorophyll content [99] and may
Remote Sens. 2021, 13, 705 20 of 26
influence its prediction from leaf spectral signal. The sampling year 2018 was rather dry
and the water content in needles exhibited the opposite (increasing) trend towards older
needles in comparison to previous season 2017 (Appendix A: Figure A2). We hypothesize
that lower water content in 1st year needles could also negatively influence chlorophyll
prediction from NDVI and NDVIred edge . The effect of leaf water content on relationship
among vegetation indices and leaf-level functional traits was observed in crops [100,101],
however not yet confirmed for conifers.
In addition, this study’s results showed that first year needles exhibited the lowest
chlorophyll contents (31–35 ug.cm2 , Table 4), and it can be hypothesized that such values,
in combination with coniferous canopy structure, may be below the detection limit of
the Parrot Sequoia multispectral camera. In a maize field case study [102], the authors
concluded that hemispherical-conical reflectance factors, NDVI, and the chlorophyll red-
edge index derived from the Sequoia sensor exhibit bias for high and low reflective surfaces.
In comparison to broadleaf trees, conifer canopy NIR reflectance is generally lower [103]
due to needle clustering within shoots and self-shading [104,105], and it is speculated that
the needle photosynthetic pigment contents of the first year needles were, in this case, too
low to be resolved by the Parrot Sequoia multispectral camera, which has limited spectral
resolution and sensitivity as compared to hyperspectral sensors.
In this study, the crown light condition showed to be much less important than the
needle age selection as a ground truth. Surprisingly the highest R2 for both VIs was
achieved when using the less sunlit, lower part of the crown (Scenario 3), followed by
Scenario 1 (full crown) and then Scenario 2 (the more sun-lit top part of the crown). It
seems that, at such a high spatial resolution, which was achieved when using the UAV
platform, the tree structure and needle/branch position can cause these differences. The
tree branches of the more sun-lit top part of the crown have a different position; they are
shorter and more pointed up, while the less sunlit lower part of the crown has wider and
flatter branches, thus the needles have a better position regarding the Sun, flight, and sensor
geometry, as shown in Figure A3 (Appendix A). Additionally, the more sunlit top part of
the crown is presented by a higher percentage of the first year needles, which were found to
be problematic for the reasons that are discussed above. To date, the effect of heterogeneous
light conditions within the crown on UAV-based leaf traits modelling was tested on broad
leaved apple and pear trees with similar result to those in this study: the full canopy spectra
provided, in some cases, more accurate models than only sunlit pixels [106]. The authors
suggest that including the signal from the whole crown results in a bigger sample size,
which may lead to model improvement. However, it can be concluded that very little
is known about this issue and a complex study on reflectance variations regarding the
tree/crown structure, needle configuration, and light conditions is still needed. It is also
important to emphasize that no study was found to directly compare these results that were
conducted on coniferous forest stands. This shows that building quantitative approaches
employing Multi-spectral Unmanned Aerial Vehicle (UAV) images is still challenging for
coniferous forests, in particular, those of natural origin. The number of the trees studied
in this research was limited (18 trees). Follow up research is needed that would allow for
extending the number of sampled trees to obtain generalizable results, especially for the
regression models.
5. Conclusions
The results show that there is a big potential in using UAVs together with affordable
multispectral cameras as a platform for monitoring forest status at a local scale, however,
at high resolution. Tree crown delineation and derivation of other parameters, such as
tree top and height, which was based on the Canopy Height Model (CHM) obtained
from two data sources—digital surface model derived from the Parrot Sequoia camera
multispectral images (Seq DSM) and high resolution Digital Elevation Model (DEM: DMR
5G), corresponded well with the in-situ data and it was satisfactory for the purposes of this
study. The results of the conducted statistical analysis show that the two tested VIs (NDVI
Remote Sens. 2021, 13, 705 21 of 26
and NDVIred edge ) have the potential to assess photosynthetic pigments in Norway spruce
forests at a semi-quantitative level; however, the selection of needle-age as a ground truth
was revealed to be a very important factor. The only usable results were obtained for linear
models when using the second year needle pigment contents as a ground truth.
On the other hand, the illumination conditions of the crown showed to have very
Remote Sens. 2021, 13, 705 little effect on the model’s validity, whereas slightly better results were obtained 22 of 28
when
assessing the less sunlit lower part of the crown, which is characterized by wider and more
flat branches. When compared to the whole crown Scenario, the improvement was very
small,
However, and itthis
is proposed that the
effect might whole
have part ofimpact
a bigger the crown be used
on data withforasimplicity.
very highHowever,
spectral
this
resolution (e.g., hyperspectral data), and further systematic researchresolution
effect might have a bigger impact on data with a very high spectral (e.g.,
on reflectance
hyperspectral data), and further systematic research on reflectance variations
variations regarding the tree/crown structure, needle configuration, and light conditions regarding
the tree/crown
is still needed. No structure,
study wasneedle configuration,
found, in which itand waslight conditions
possible is still
to directly needed.these
compare No
study
resultswas found, inonwhich
conducted it was (Norway
coniferous possible to directly
spruce) compare
forest these
stands; results
this showsconducted
that thereonis
coniferous (Norway spruce) forest stands; this shows that there is also a
also a further need for studies dealing with a quantitative estimation of the biochemicalfurther need for
studies
variables dealing with a quantitative
of coniferous forests when estimation
employing of the biochemical
spectral data thatvariables of coniferous
were acquired at the
forests when employing spectral data
UAV platform at a very high spatial resolution. that were acquired at the UAV platform at a very
high spatial resolution.
Author Contributions: Conceptualization, V.K.-S.; methodology, V.K.-S.; software, L.K.;
Author Contributions:
validation, Conceptualization,
F.O., Z.L.; formal analysis, L.K., V.K.-S.;
J.J., Z.L.;methodology,
investigation,V.K.-S.;
V.K.-S.,software, L.K.;
F.O., Z.L.; validation,
resources, J.J.,
F.O.,
F.O., Z.L.; writing—original draft preparation, L.K., V.K.-S., Z.L., J.J., F.O.; reviewing and F.O.,
Z.L.; formal analysis, L.K., J.J., Z.L.; investigation, V.K.-S., F.O., Z.L.; resources, J.J., Z.L.;
editing of
writing—original
the first manuscriptdraft preparation,
version, V.K.-S.; L.K., V.K.-S.,
revision: Z.L.,visualization,
V.K.-S., J.J., F.O.; reviewing and
L.K., J.J.; editing of the
supervision, first
V.K.-S.;
manuscript version, V.K.-S.;
project administration, revision:
V.K.-S.; V.K.-S.,
All authors havevisualization,
read and agreedL.K., J.J.; supervision,
to the publishedV.K.-S.;
versionproject
of the
administration,
manuscript. V.K.-S.; All authors have read and agreed to the published version of the manuscript.
Funding: This
Funding: This research
research was
was funded
funded byby Czech
Czech Science
Science Foundation
Foundation (grant
(grant number
number 17-05743S)
17-05743S) and
and
Czech Geological Survey, grant no. 310360.
Czech Geological Survey, grant no. 310360.
Institutional
Institutional Review
Review Board Statement: Not
Board Statement: Not applicable.
applicable.
Informed
Informed Consent Statement:Informed
ConsentStatement: Informedconsent was
consent obtained
was fromfrom
obtained all subjects involved
all subjects in the in
involved study.
the
study.
Data Availability Statement: The data presented in this study are available on reasonable request
from
Data the corresponding
Availability author.The data presented in this study are available on reasonable request
Statement:
from the corresponding
Acknowledgments: author.
We thank to Miroslav Barták for helping us with the chlorophyll analyses.
Acknowledgments: We thank to Miroslav Barták for helping us with the chlorophyll analyses.
Conflicts of Interest: The authors declare no conflict of interest.
Conflicts of Interest: The authors declare no conflict of interest.
Appendix A
Appendix A
FigureA1.
Figure A1.In-situ
In-situ ground
ground truth - biochemically
truth-biochemically assessed
assessed totaltotal chlorophyll
chlorophyll content
content in different
in different needle
needle age classes
age classes of studied
of studied
trees. In most cases chlorophyll content increased from 1st to 2nd year needles and usually the pigment further accumulated
trees. In most cases chlorophyll content increased from 1st to 2nd year needles and usually the pigment further accumulated
ininolder
olderneedles.
needles.
Remote Sens.Sens.
Remote 13, 705
2021,2021, 13, 705 22 of 26 23 of 28
2021, 13, 705 23 of 28
Figure A2. Selected biophysical traits ofbiophysical

Figure A2. Selected needles attraits
LYSofand PLBatsites
needles in 2017
LYS and PLB (year
sites inbefore sampling)
2017 (year and 2018 (year o
before sampling)
sampling and UAV data acquisition). (a,b) Leaf mass per area; (c,d) Needle chlorophyll content; (e,f)
and 2018 (year of sampling and UAV data acquisition). (a,b) Leaf mass per area; (c,d) NeedleNeedle water conten
e A2. Selected biophysical
per area traits
basis. of needles
NAC at LYS
ageand
=chlorophyll
needle PLB
class: sites in
1—first 2017
year (year before
needles, 2—secondsampling)
year and 2018
needles, (year years
4—four of and older needles.
content; (e,f) Needle water content per area basis. NAC = needle age class: 1—first year
ling and UAV data acquisition). (a,b) Leaf mass per area; (c,d) Needle chlorophyll content; (e,f) Needle water content
needles, 2—second year year
rea basis. NAC = needle age class: 1—first year needles, 2—second needles, 4—four
needles, yearsyears
4—four and older needles.
and older needles.
Figure A3. 3D representation of a solitary growing coniferous tree with a Norway spruce-like
Figure
Figure A3. 3D representation of a3D
A3.
crown representation
solitary Sideof
growing
architecture. a solitary
coniferous
(a) growing
tree
and nadir with coniferous
view a(b). tree with a Norway spruce-like crown
Norway spruce-like
architecture. Side (a) and
crown architecture. Side (a) and nadir view (b). nadir view (b).
Remote Sens. 2021, 13, 705 23 of 26
References
1. Potapov, P.; Yaroshenko, A.; Turubanova, S.; Dubinin, M.; Laestadius, L.; Thies, C.; Aksenov, D.; Egorov, A.; Yesipova, Y.;
Glushkov, I. Mapping the World’s Intact Forest Landscapes by Remote Sensing. Ecol. Soc. 2008, 13, 51. [CrossRef]
2. Boyd, D.S.; Danson, F.M. Satellite Remote Sensing of Forest Resources: Three Decades of Research Development. Prog. Phys.
Geogr. 2005, 29, 1–26. [CrossRef]
3. De Vries, W.; Vel, E.; Reinds, G.J.; Deelstra, H.; Klap, J.M.; Leeters, E.; Hendriks, C.M.A.; Kerkvoorden, M.; Landmann, G.;
Herkendell, J. Intensive Monitoring of Forest Ecosystems in Europe: 1. Objectives, Set-up and Evaluation Strategy. For. Ecol.
Manag. 2003, 174, 77–95. [CrossRef]
4. Pause, M.; Schweitzer, C.; Rosenthal, M. In Situ/Remote Sensing Integration to Assess Forest Health—A Review. Remote Sens.
2016, 8, 471. [CrossRef]
5. Žibret, G.; Kopačková, V. Comparison of Two Methods for Indirect Measurement of Atmospheric Dust Deposition: Street-Dust
Composition and Vegetation-Health Status Derived from Hyperspectral Image Data. Ambio 2019, 48, 423–435. [CrossRef]
6. Hruška, J.; Krám, P. Modelling Long-Term Changes in Stream Water and Soil Chemistry in Catchments with Contrasting
Vulnerability to Acidification (Lysina and Pluhuv Bor, Czech Republic). Hydrol. Earth Syst. Sci. 2003, 7, 525–539. [CrossRef]
7. Švik, M.; Oulehle, F.; Krám, P.; Janoutová, R.; Tajovská, K.; Homolová, L. Landsat-Based Indices Reveal Consistent Recovery of
Forested Stream Catchments from Acid Deposition. Remote Sens. 2020, 12, 1944. [CrossRef]
8. Fottová, D.; Skořepová, I. Changes in mass element fluxes and their importance for critical loads: GEOMON network, Czech
Republic. In Biogeochemical Investigations at Watershed, Landscape, and Regional Scales; Springer: Berlin/Heidelberg, Germany, 1998;
pp. 365–376.
9. Fottová, D. Trends in Sulphur and Nitrogen Deposition Fluxes in the GEOMON Network, Czech Republic, between 1994 and
2000. Water. Air. Soil Pollut. 2003, 150, 73–87. [CrossRef]
10. Oulehle, F.; Chuman, T.; Hruška, J.; Krám, P.; McDowell, W.H.; Myška, O.; Navrátil, T.; Tesař, M. Recovery from Acidification
Alters Concentrations and Fluxes of Solutes from Czech Catchments. Biogeochemistry 2017, 132, 251–272. [CrossRef]
11. Krám, P.; Hruška, J.; Shanley, J.B. Streamwater Chemistry in Three Contrasting Monolithologic Czech Catchments. Appl. Geochem.
2012, 27, 1854–1863. [CrossRef]
12. Navrátil, T.; Kurz, D.; Krám, P.; Hofmeister, J.; Hruška, J. Acidification and Recovery of Soil at a Heavily Impacted Forest
Catchment (Lysina, Czech Republic)—SAFE Modeling and Field Results. Ecol. Model. 2007, 205, 464–474. [CrossRef]
13. Oulehle, F.; McDowell, W.H.; Aitkenhead-Peterson, J.A.; Krám, P.; Hruška, J.; Navrátil, T.; Buzek, F.; Fottová, D. Long-Term
Trends in Stream Nitrate Concentrations and Losses across Watersheds Undergoing Recovery from Acidification in the Czech
Republic. Ecosystems 2008, 11, 410–425. [CrossRef]
14. dos Santos, A.A.; Junior, J.M.; Araújo, M.S.; Di Martini, D.R.; Tetila, E.C.; Siqueira, H.L.; Aoki, C.; Eltner, A.; Matsubara, E.T.;
Pistori, H. Assessment of CNN-Based Methods for Individual Tree Detection on Images Captured by RGB Cameras Attached to
UAVs. Sensors 2019, 19, 3595. [CrossRef]
15. Fujimoto, A.; Haga, C.; Matsui, T.; Machimura, T.; Hayashi, K.; Sugita, S.; Takagi, H. An End to End Process Development for
UAV-SfM Based Forest Monitoring: Individual Tree Detection, Species Classification and Carbon Dynamics Simulation. Forests
2019, 10, 680. [CrossRef]
16. Fayad, I.; Baghdadi, N.; Guitet, S.; Bailly, J.-S.; Hérault, B.; Gond, V.; El Hajj, M.; Minh, D.H.T. Aboveground Biomass Mapping in
French Guiana by Combining Remote Sensing, Forest Inventories and Environmental Data. Int. J. Appl. Earth Obs. Geoinform.
2016, 52, 502–514. [CrossRef]
17. White, J.C.; Coops, N.C.; Wulder, M.A.; Vastaranta, M.; Hilker, T.; Tompalski, P. Remote Sensing Technologies for Enhancing
Forest Inventories: A Review. Can. J. Remote Sens. 2016, 42, 619–641. [CrossRef]
18. Masek, J.G.; Hayes, D.J.; Hughes, M.J.; Healey, S.P.; Turner, D.P. The Role of Remote Sensing in Process-Scaling Studies of
Managed Forest Ecosystems. For. Ecol. Manag. 2015, 355, 109–123. [CrossRef]
19. Lausch, A.; Heurich, M.; Gordalla, D.; Dobner, H.-J.; Gwillym-Margianto, S.; Salbach, C. Forecasting Potential Bark Beetle
Outbreaks Based on Spruce Forest Vitality Using Hyperspectral Remote-Sensing Techniques at Different Scales. For. Ecol. Manag.
2013, 308, 76–89. [CrossRef]
20. Halme, E.; Pellikka, P.; Mõttus, M. Utility of Hyperspectral Compared to Multispectral Remote Sensing Data in Estimating Forest
Biomass and Structure Variables in Finnish Boreal Forest. Int. J. Appl. Earth Obs. Geoinform. 2019, 83, 101942. [CrossRef]
21. Jarocińska, A.; Białczak, M.; Sławik, Ł. Application of Aerial Hyperspectral Images in Monitoring Tree Biophysical Parameters in
Urban Areas. Misc. Geogr. 2018, 22, 56–62. [CrossRef]
22. Kopačková, V.; Mišurec, J.; Lhotáková, Z.; Oulehle, F.; Albrechtová, J. Using Multi-Date High Spectral Resolution Data to Assess
the Physiological Status of Macroscopically Undamaged Foliage on a Regional Scale. Int. J. Appl. Earth Obs. Geoinform. 2014, 27,
169–186. [CrossRef]
23. Machala, M.; Zejdová, L. Forest Mapping Through Object-Based Image Analysis of Multispectral and LiDAR Aerial Data. Eur. J.
Remote Sens. 2014, 47, 117–131. [CrossRef]
24. Mišurec, J.; Kopačková, V.; Lhotáková, Z.; Campbell, P.; Albrechtová, J. Detection of Spatio-Temporal Changes of Norway
Spruce Forest Stands in Ore Mountains Using Landsat Time Series and Airborne Hyperspectral Imagery. Remote Sens. 2016, 8,
92. [CrossRef]
Remote Sens. 2021, 13, 705 24 of 26
25. Lin, Y.; Jiang, M.; Yao, Y.; Zhang, L.; Lin, J. Use of UAV Oblique Imaging for the Detection of Individual Trees in Residential
Environments. Urban For. Urban Green. 2015, 14, 404–412. [CrossRef]
26. Klein Hentz, Â.M.; Corte, A.P.D.; Péllico Netto, S.; Strager, M.P.; Schoeninger, E.R. Treedetection: Automatic Tree Detection Using
UAV-Based Data. Floresta 2018, 48, 393. [CrossRef]
27. Candiago, S.; Remondino, F.; De Giglio, M.; Dubbini, M.; Gattelli, M. Evaluating Multispectral Images and Vegetation Indices for
Precision Farming Applications from UAV Images. Remote Sens. 2015, 7, 4026–4047. [CrossRef]
28. Sona, G.; Passoni, D.; Pinto, L.; Pagliari, D.; Masseroni, D.; Ortuani, B.; Facchi, A. UAV Multispectral Survey to Map Soil and Crop
for Precision Farming Applications. ISPRS Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2016, 1023–1029. [CrossRef]
29. Albetis, J.; Duthoit, S.; Guttler, F.; Jacquin, A.; Goulard, M.; Poilvé, H.; Féret, J.-B.; Dedieu, G. Detection of Flavescence Dorée
Grapevine Disease Using Unmanned Aerial Vehicle (UAV) Multispectral Imagery. Remote Sens. 2017, 9, 308. [CrossRef]
30. Su, J.; Liu, C.; Hu, X.; Xu, X.; Guo, L.; Chen, W.-H. Spatio-Temporal Monitoring of Wheat Yellow Rust Using UAV Multispectral
Imagery. Comput. Electron. Agric. 2019, 167, 105035. [CrossRef]
31. Singhal, G.; Bansod, B.; Mathew, L.; Goswami, J.; Choudhury, B.U.; Raju, P.L.N. Chlorophyll Estimation Using Multi-Spectral
Unmanned Aerial System Based on Machine Learning Techniques. Remote Sens. Appl. Soc. Environ. 2019, 15, 100235. [CrossRef]
32. Guo, Y.; Wang, H.; Wu, Z.; Wang, S.; Sun, H.; Senthilnath, J.; Wang, J.; Robin Bryant, C.; Fu, Y. Modified Red Blue Vegetation
Index for Chlorophyll Estimation and Yield Prediction of Maize from Visible Images Captured by UAV. Sensors 2020, 20, 5055.
[CrossRef] [PubMed]
33. Baluja, J.; Diago, M.P.; Balda, P.; Zorer, R.; Meggio, F.; Morales, F.; Tardaguila, J. Assessment of Vineyard Water Status Variability
by Thermal and Multispectral Imagery Using an Unmanned Aerial Vehicle (UAV). Irrig. Sci. 2012, 30, 511–522. [CrossRef]
34. Turner, D.; Lucieer, A.; Watson, C. Development of an Unmanned Aerial Vehicle (UAV) for Hyper Resolution Vineyard Mapping
Based on Visible, Multispectral, and Thermal Imagery. In Proceedings of the 34th International Symposium on Remote Sensing of
Environment, Sydney, NSW, Australia, 10–15 April 2011; p. 4.
35. Zarco-Tejada, P.J.; Guillén-Climent, M.L.; Hernández-Clemente, R.; Catalina, A.; González, M.R.; Martín, P. Estimating Leaf
Carotenoid Content in Vineyards Using High Resolution Hyperspectral Imagery Acquired from an Unmanned Aerial Vehicle
(UAV). Agric. For. Meteorol. 2013, 171–172, 281–294. [CrossRef]
36. Belmonte, A.; Sankey, T.; Biederman, J.A.; Bradford, J.; Goetz, S.J.; Kolb, T.; Woolley, T. UAV-derived Estimates of Forest Structure
to Inform Ponderosa Pine Forest Restoration. Remote Sens. Ecol. Conserv. 2020, 6, 181–197. [CrossRef]
37. D’Odorico, P.; Besik, A.; Wong, C.Y.S.; Isabel, N.; Ensminger, I. High-throughput Drone-based Remote Sensing Reliably Tracks
Phenology in Thousands of Conifer Seedlings. New Phytol. 2020, 226, 1667–1681. [CrossRef] [PubMed]
38. Dash, J.P.; Watt, M.S.; Pearse, G.D.; Heaphy, M.; Dungey, H.S. Assessing Very High Resolution UAV Imagery for Monitoring
Forest Health during a Simulated Disease Outbreak. ISPRS J. Photogramm. Remote Sens. 2017, 131, 1–14. [CrossRef]
39. Chianucci, F.; Disperati, L.; Guzzi, D.; Bianchini, D.; Nardino, V.; Lastri, C.; Rindinella, A.; Corona, P. Estimation of Canopy
Attributes in Beech Forests Using True Colour Digital Images from a Small Fixed-Wing UAV. Int. J. Appl. Earth Obs. Geoinform.
2016, 47, 60–68. [CrossRef]
40. Dash, J.; Pearse, G.; Watt, M. UAV Multispectral Imagery Can Complement Satellite Data for Monitoring Forest Health. Remote
Sens. 2018, 10, 1216. [CrossRef]
41. Minařík, R.; Langhammer, J. Use of a Multispectral UAV Photogrammetry for Detection and Tracking of Forest Disturbance
Dynamics. ISPRS Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2016, 711–718. [CrossRef]
42. Franklin, S.E. Pixel- and Object-Based Multispectral Classification of Forest Tree Species from Small Unmanned Aerial Vehicles. J.
Unmanned Veh. Syst. 2018, 6, 195–211. [CrossRef]
43. Näsi, R.; Honkavaara, E.; Lyytikäinen-Saarenmaa, P.; Blomqvist, M.; Litkey, P.; Hakala, T.; Viljanen, N.; Kantola, T.; Tanhuanpää, T.;
Holopainen, M. Using UAV-Based Photogrammetry and Hyperspectral Imaging for Mapping Bark Beetle Damage at Tree-Level.
Remote Sens. 2015, 7, 15467–15493. [CrossRef]
44. Berveglieri, A.; Tommaselli, A.M.G. Exterior Orientation of Hyperspectral Frame Images Collected with UAV for Forest Applica-
tions. ISPRS Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2016, 40, 45–50. [CrossRef]
45. Gallardo-Salazar, J.L.; Pompa-García, M. Detecting Individual Tree Attributes and Multispectral Indices Using Unmanned Aerial
Vehicles: Applications in a Pine Clonal Orchard. Remote Sens. 2020, 12, 4144. [CrossRef]
46. Lim, Y.S.; La, P.H.; Park, J.S.; Lee, M.H.; Pyeon, M.W.; Kim, J.-I. Calculation of Tree Height and Canopy Crown from Drone Images
Using Segmentation. J. Kor. Soc. Survey. Geodesy Photogram. Cartogr. 2015, 33, 605–614. [CrossRef]
47. Díaz-Varela, R.; de la Rosa, R.; León, L.; Zarco-Tejada, P. High-Resolution Airborne UAV Imagery to Assess Olive Tree Crown
Parameters Using 3D Photo Reconstruction: Application in Breeding Trials. Remote Sens. 2015, 7, 4213–4232. [CrossRef]
48. Wu, X.; Shen, X.; Cao, L.; Wang, G.; Cao, F. Assessment of Individual Tree Detection and Canopy Cover Estimation Using
Unmanned Aerial Vehicle Based Light Detection and Ranging (UAV-LiDAR) Data in Planted Forests. Remote Sens. 2019, 11,
908. [CrossRef]
49. Panagiotidis, D.; Abdollahnejad, A.; Surový, P.; Chiteculo, V. Determining Tree Height and Crown Diameter from High-Resolution
UAV Imagery. Int. J. Remote Sens. 2017, 38, 2392–2410. [CrossRef]
50. Jaakkola, A.; Hyyppä, J.; Yu, X.; Kukko, A.; Kaartinen, H.; Liang, X.; Hyyppä, H.; Wang, Y. Autonomous Collection of Forest Field
Reference—The Outlook and a First Step with UAV Laser Scanning. Remote Sens. 2017, 9, 785. [CrossRef]
Remote Sens. 2021, 13, 705 25 of 26
51. Silva, C.A.; Crookston, N.L.; Hudak, A.T.; Vierling, L.A.; Klauberg, C.; Cardil, A. RLiDAR: LiDAR Data Processing and Visualization;
The R Foundation: Vienna, Austria, 2017.
52. Cardil, A.; Otsu, K.; Pla, M.; Silva, C.A.; Brotons, L. Quantifying Pine Processionary Moth Defoliation in a Pine-Oak Mixed Forest
Using Unmanned Aerial Systems and Multispectral Imagery. PLoS ONE 2019, 14, e0213027. [CrossRef]
53. Zaforemska, A.; Xiao, W.; Gaulton, R. Individual Tree Detection from UAV LIDAR Data in a Mixed Species Woodland. ISPRS Int.
Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2019, 657–663. [CrossRef]
54. Li, J.; Yang, B.; Cong, Y.; Cao, L.; Fu, X.; Dong, Z. 3D Forest Mapping Using A Low-Cost UAV Laser Scanning System: Investigation
and Comparison. Remote Sens. 2019, 11, 717. [CrossRef]
55. Kuželka, K.; Slavík, M.; Surový, P. Very High Density Point Clouds from UAV Laser Scanning for Automatic Tree Stem Detection
and Direct Diameter Measurement. Remote Sens. 2020, 12, 1236. [CrossRef]
56. Picos, J.; Bastos, G.; Míguez, D.; Alonso, L.; Armesto, J. Individual Tree Detection in a Eucalyptus Plantation Using Unmanned
Aerial Vehicle (UAV)-LiDAR. Remote Sens. 2020, 12, 885. [CrossRef]
57. Pulido, D.; Salas, J.; Rös, M.; Puettmann, K.; Karaman, S. Assessment of Tree Detection Methods in Multispectral Aerial Images.
Remote Sens. 2020, 12, 2379. [CrossRef]
58. Ples, oianu, A.-I.; Stupariu, M.-S.; S, andric, I.; Pătru-Stupariu, I.; Drăgut, , L. Individual Tree-Crown Detection and Species Clas-
sification in Very High-Resolution Remote Sensing Imagery Using a Deep Learning Ensemble Model. Remote Sens. 2020, 12,
2426. [CrossRef]
59. Fromm, M.; Schubert, M.; Castilla, G.; Linke, J.; McDermid, G. Automated Detection of Conifer Seedlings in Drone Imagery
Using Convolutional Neural Networks. Remote Sens. 2019, 11, 2585. [CrossRef]
60. Lhotáková, Z.; Kopačková-Strnadová, V.; Oulehle, F.; Homolová, L.; Neuwirthová, E.; Švik, M.; Janoutová, R.; Albrechtová, J.
Foliage Biophysical Trait Prediction from Laboratory Spectra in Norway Spruce Is More Affected by Needle Age Than by Site
Soil Conditions. Remote Sens. 2021, 13, 391. [CrossRef]
61. ČÚZK: Geoportál. Available online: https://geoportal.cuzk.cz/(S(5rmwjwtpueumnrsepeeaarwk))/Default.aspx?head_tab=
sekce-00-gp&mode=TextMeta&text=uvod_uvod&menu=01&news=yes&UvodniStrana=yes (accessed on 31 December 2020).
62. Novotný, J.; Navrátilová, B.; Janoutová, R.; Oulehle, F.; Homolová, L. Influence of Site-Specific Conditions on Estimation of Forest
above Ground Biomass from Airborne Laser Scanning. Forests 2020, 11, 268. [CrossRef]
63. Porra, R.J.; Thompson, W.A.; Kriedemann, P.E. Determination of Accurate Extinction Coefficients and Simultaneous Equations
for Assaying Chlorophylls a and b Extracted with Four Different Solvents: Verification of the Concentration of Chlorophyll
Standards by Atomic Absorption Spectroscopy. Biochim. Biophys. Acta BBA Bioenerg. 1989, 975, 384–394. [CrossRef]
64. Wellburn, A.R. The Spectral Determination of Chlorophylls a and b, as Well as Total Carotenoids, Using Various Solvents with
Spectrophotometers of Different Resolution. J. Plant Physiol. 1994, 144, 307–313. [CrossRef]
65. DJI—Official Website. Available online: https://www.dji.com (accessed on 30 December 2020).
66. Phantom 4—DJI. Available online: https://www.dji.com/phantom-4 (accessed on 30 December 2020).
67. SenseFly—SenseFly—The Professional’s Mapping Drone. Available online: https://www.sensefly.com/ (accessed on
30 December 2020).
68. Prusa3D—Open-Source 3D Printers from Josef Prusa. Available online: https://www.prusa3d.com/ (accessed on
30 December 2020).
69. Litchi for DJI Mavic/Phantom/Inspire/Spark. Available online: https://flylitchi.com/ (accessed on 31 December 2020).
70. Best Practices: Collecting Data with MicaSense Sensors. Available online: https://support.micasense.com/hc/en-us/articles/22
4893167-Best-practices-Collecting-Data-with-MicaSense-Sensors (accessed on 31 December 2020).
71. Agisoft Metashape. Available online: https://www.agisoft.com/ (accessed on 31 December 2020).
72. Westoby, M.J.; Brasington, J.; Glasser, N.F.; Hambrey, M.; Reynolds, J.M. ‘Structure-from-Motion’ Photogrammetry: A Low-Cost,
Effective Tool for Geoscience Applications. Geomorphology 2012, 179, 300–314. [CrossRef]
73. Tutorial (Intermediate Level): Radiometric Calibration Using Reflectance Panelsin PhotoScan Professional 1.4. Available online:
https://www.agisoft.com/pdf/PS_1.4_(IL)_Refelctance_Calibration.pdf (accessed on 20 December 2020).
74. Zhao, K.; Popescu, S. Hierarchical Watershed Segmentation of Canopy Height Model for Multi-Scale Forest Inventory. In
Proceedings of the ISPRS working group “Laser Scanning 2007 and SilviLaser 2007”, Espoo, Finland, 12–14 September 2007;
pp. 436–440.
75. Hubacek, M.; Kovarik, V.; Kratochvil, V. Analysis of Influence of Terrain Relief Roughness on DEM Accuracy Generated from
LIDAR in the Czech Republic Territory. In Proceedings of the ISPRS—International Archives of the Photogrammetry, Remote
Sensing and Spatial Information Sciences, Prague, Czech Republic, 12–19 July 2016; Copernicus GmbH: Göttingen, Germany;
Volume XLI-B4, pp. 25–30.
76. How Focal Statistics Works—Help|ArcGIS for Desktop. Available online: https://desktop.arcgis.com/en/arcmap/10.3/tools/
spatial-analyst-toolbox/how-focal-statistics-works.htm (accessed on 31 December 2020).
77. Xu, N.; Tian, J.; Tian, Q.; Xu, K.; Tang, S. Analysis of Vegetation Red Edge with Different Illuminated/Shaded Canopy Proportions
and to Construct Normalized Difference Canopy Shadow Index. Remote Sens. 2019, 11, 1192. [CrossRef]
78. Rouse, J.W.; Hass, R.H.; Schell, J.A.; Deering, D.W.; Harlan, J.C. Monitoring the Vernal Advancement and Retrogradation (Green Wave
Effect) of Natural Vegetation.; RSC 1978-4; Texas A & M University: College Station, TX, USA, 1974.
Remote Sens. 2021, 13, 705 26 of 26
79. Sharma, L.; Bu, H.; Denton, A.; Franzen, D. Active-Optical Sensors Using Red NDVI Compared to Red Edge NDVI for Prediction
of Corn Grain Yield in North Dakota, U.S.A. Sensors 2015, 15, 27832–27853. [CrossRef] [PubMed]
80. Gitelson, A.; Merzlyak, M.N. Spectral Reflectance Changes Associated with Autumn Senescence of Aesculus Hippocastanum
L. and Acer Platanoides L. Leaves. Spectral Features and Relation to Chlorophyll Estimation. J. Plant Physiol. 1994, 143,
286–292. [CrossRef]
81. Sims, D.A.; Gamon, J.A. Relationships between Leaf Pigment Content and Spectral Reflectance across a Wide Range of Species,
Leaf Structures and Developmental Stages. Remote Sens. Environ. 2002, 81, 337–354. [CrossRef]
82. Chavez, P.S., Jr.; Kwarteng, A.Y. Extracting Spectral Contrast in Landsat Thematic Mapper Image Data Using Selective Principal
Component Analysis. Photogramm. Eng. Remote Sens. 1989, 55, 10.
83. Mather, P.M.; Koch, M. Computer Processing of Remotely-Sensed Images: An Introduction, 4th ed.; Wiley-Blackwell: Chichester, UK;
Hoboken, NJ, USA, 2011; ISBN 978-0-470-74239-6.
84. Draper, N.R.; Smith, H. Applied Regression Analysis. In Wiley Series in Probability and Statistics, 1st ed.; Wiley: Hoboken, NJ, USA,
1998; ISBN 978-0-471-17082-2.
85. Homolová, L.; Lukeš, P.; Malenovský, Z.; Lhotáková, Z.; Kaplan, V.; Hanuš, J. Measurement Methods and Variability Assessment
of the Norway Spruce Total Leaf Area: Implications for Remote Sensing. Trees 2013, 27, 111–121. [CrossRef]
86. Krám, P.; Oulehle, F.; Štědrá, V.; Hruška, J.; Shanley, J.B.; Minocha, R.; Traister, E. Geoecology of a Forest Watershed Underlain by
Serpentine in Central Europe. Northeast. Nat. 2009, 16, 309–328. [CrossRef]
87. O’Neill, A.L.; Kupiec, J.A.; Curran, P.J. Biochemical and Reflectance Variation throughout a Sitka Spruce Canopy. Remote Sens.
Environ. 2002, 80, 134–142. [CrossRef]
88. Hovi, A.; Raitio, P.; Rautiainen, M. A Spectral Analysis of 25 Boreal Tree Species. Silva Fenn. 2017, 51. [CrossRef]
89. Rautiainen, M.; Lukeš, P.; Homolová, L.; Hovi, A.; Pisek, J.; Mõttus, M. Spectral Properties of Coniferous Forests: A Review of In
Situ and Laboratory Measurements. Remote Sens. 2018, 10, 207. [CrossRef]
90. Wu, Q.; Song, C.; Song, J.; Wang, J.; Chen, S.; Yu, B. Impacts of Leaf Age on Canopy Spectral Signature Variation in Evergreen
Chinese Fir Forests. Remote Sens. 2018, 10, 262. [CrossRef]
91. Rogers, S.R.; Manning, I.; Livingstone, W. Comparing the Spatial Accuracy of Digital Surface Models from Four Unoccupied
Aerial Systems: Photogrammetry Versus LiDAR. Remote Sens. 2020, 12, 2806. [CrossRef]
92. Zhang, L.; Sun, X.; Wu, T.; Zhang, H. An analysis of shadow effects on spectral vegetation indexes using a ground-based imaging
spectrometer. IEEE Geosci. Remote Sens. Lett. 2015, 12, 2188–2192. [CrossRef]
93. Rock, B.N.; Hoshizaki, T.; Miller, J.R. Comparison of in Situ and Airborne Spectral Measurements of the Blue Shift Associated
with Forest Decline. Remote Sens. Environ. 1988, 24, 109–127. [CrossRef]
94. Campbell, P.K.E.; Rock, B.N.; Martin, M.E.; Neefus, C.D.; Irons, J.R.; Middleton, E.M.; Albrechtova, J. Detection of Initial Damage
in Norway Spruce Canopies Using Hyperspectral Airborne Data. Int. J. Remote Sens. 2004, 25, 5557–5584. [CrossRef]
95. de Tomás Marín, S.; Novák, M.; Klančnik, K.; Gaberščik, A. Spectral Signatures of Conifer Needles Mainly Depend on Their
Physical Traits. Pol. J. Ecol. 2016, 64, 1–13. [CrossRef]
96. Misurec, J.; Kopacková, V.; Lhotakova, Z.; Albrechtova, J.; Hanus, J.; Weyermann, J.; Entcheva-Campbell, P. Utilization of Hyper-
spectral Image Optical Indices to Assess the Norway Spruce Forest Health Status. J. Appl. Remote Sens. 2012, 6, 063545. [CrossRef]
97. Hernández-Clemente, R.; Navarro-Cerrillo, R.M.; Zarco-Tejada, P.J. Carotenoid Content Estimation in a Heterogeneous Conifer
Forest Using Narrow-Band Indices and PROSPECT+DART Simulations. Remote Sens. Environ. 2012, 127, 298–315. [CrossRef]
98. Kováč, D.; Malenovský, Z.; Urban, O.; Špunda, V.; Kalina, J.; Ač, A.; Kaplan, V.; Hanuš, J. Response of Green Reflectance
Continuum Removal Index to the Xanthophyll De-Epoxidation Cycle in Norway Spruce Needles. J. Exp. Bot. 2013, 64, 1817–1827.
[CrossRef] [PubMed]
99. Kokaly, R.F.; Asner, G.P.; Ollinger, S.V.; Martin, M.E.; Wessman, C.A. Characterizing Canopy Biochemistry from Imaging
Spectroscopy and Its Application to Ecosystem Studies. Remote Sens. Environ. 2009, 113, S78–S91. [CrossRef]
100. Klem, K.; Záhora, J.; Zemek, F.; Trunda, P.; Tůma, I.; Novotná, K.; Hodaňová, K.; Rapantová, B.; Hanuš, J.; Vaříková, J.; et al.
Interactive effects of water deficit and nitrogen nutrition on winter wheat. Remote sensing methods for their detection. Agricult.
Water Manag. 2018, 210, 171–184. [CrossRef]
101. Schlemmer, M.R.; Francis, D.D.; Shanahan, J.F.; Schepers, J.S. Remotely measuring chlorophyll content in corn leaves with
differing nitrogen levels and relative water content. Agron. J. 2005, 97, 106–112. [CrossRef]
102. Fawcett, D.; Panigada, C.; Tagliabue, G.; Boschetti, M.; Celesti, M.; Evdokimov, A.; Biriukova, K.; Colombo, R.; Miglietta, F.;
Rascher, U.; et al. Multi-Scale Evaluation of Drone-Based Multispectral Surface Reflectance and Vegetation Indices in Operational
Conditions. Remote Sens. 2020, 12, 514. [CrossRef]
103. Ollinger, S.V. Sources of Variability in Canopy Reflectance and the Convergent Properties of Plants: Tansley Review. New Phytol.
2011, 189, 375–394. [CrossRef]
104. Cescatti, A.; Zorer, R. Structural Acclimation and Radiation Regime of Silver Fir (Abies Alba Mill.) Shoots along a Light Gradient:
Shoot Structure and Radiation Regime. Plant Cell Environ. 2003, 26, 429–442. [CrossRef]
105. Kubínová, Z.; Janáček, J.; Lhotáková, Z.; Šprtová, M.; Kubínová, L.; Albrechtová, J. Norway Spruce Needle Size and Cross Section
Shape Variability Induced by Irradiance on a Macro- and Microscale and CO2 Concentration. Trees 2018, 32, 231–244. [CrossRef]
106. Vanbrabant, Y.; Tits, L.; Delalieux, S.; Pauly, K.; Verjans, W.; Somers, B. Multitemporal Chlorophyll Mapping in Pome Fruit
Orchards from Remotely Piloted Aircraft Systems. Remote Sens. 2019, 11, 1468. [CrossRef]

Remotesensing 13 00705 v3

Uploaded by

Copyright:

Available Formats

Remotesensing 13 00705 v3

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Remotesensing 13 00705 v3

Uploaded by

Copyright:

Available Formats

remote sensing

Remote Sens. 2021, 13, 705. https://doi.org/10.3390/rs13040705 https://www.mdpi.com/journal/remotesensing

For such quantitative spectral-based approaches, which employ very high-spatial

2. Materials and Methods

Table 1. Characteristics of selected catchments: Lysina and Pluhův Bor [10].

Remote Sens. 2021, 13, 705 5 of 26

Figure 2.2. Unmanned

2.3.2. Data Acquisition

2.4. Tree Height, Crown and Top Detection

CHM = Seq DSM − DMR 5G (1)

2.5. Multispectral Data Processing

3.2. UAV Photogrammetric Products

the 20 highest trees at each stand compared to the in-situ data.

3.4. Tree Crown Illumination Classes

Remote Sens. 2021, 13, 705 15 of 28

Ground Truth Age Ground Truth Age

Table 11. Coefficients of determination

2021, 13, 705 23 of 28

Figure A2. Selected biophysical traits ofbiophysical

You might also like